Documente Academic
Documente Profesional
Documente Cultură
1 3Reference
The volumes in this series form the worlds most comprehensive reference on the
plant sciences. Composed of ten volumes, The Plant Sciences provides both
background and essential information in plant biology, exploring such topics as
genetics and genomics, molecular biology, biochemistry, growth and development,
and ecology and the environment. Available through both print and online
mediums, the online text will be continuously updated to enable the reference to
remain a useful authoritative resource for decades to come.
With broad contributions from internationally well-respected scientists in the
field, The Plant Sciences is an invaluable reference for upper-division undergraduates, graduate students, and practitioners looking for an entry into a particular
topic.
Series Titles
1. Genetics and Genomics
2. Molecular Biology
3. Biochemistry
4. Cell Biology
5. Growth and Development
6. Physiology and Function
7. Biotic Interactions
8. Ecology and the Environment
9. Evolution, Systematics and Biodiversity
10. Applications
Russell K. Monson
Editor
Editor
Russell K. Monson
School of Natural Resources and the Environment and
Laboratory for Tree Ring Research
University of Arizona
Tucson, USA
and
Professor Emeritus, Ecology and Evolutionary Biology
University of Colorado
Boulder, CO
USA
ISBN 978-1-4614-7500-2
ISBN 978-1-4614-7501-9 (eBook)
ISBN 978-1-4614-7502-6 (print and electronic bundle)
DOI 10.1007/978-1-4614-7501-9
Springer Dordrecht Heidelberg New York London
Library of Congress Control Number: 2014948194
# Springer Science+Business Media New York 2014
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of
the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or
information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed. Exempted from this legal reservation are brief excerpts
in connection with reviews or scholarly analysis or material supplied specifically for the purpose of being
entered and executed on a computer system, for exclusive use by the purchaser of the work. Duplication
of this publication or parts thereof is permitted only under the provisions of the Copyright Law of the
Publishers location, in its current version, and permission for use must always be obtained from
Springer. Permissions for use may be obtained through RightsLink at the Copyright Clearance Center.
Violations are liable to prosecution under the respective Copyright Law.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt
from the relevant protective laws and regulations and therefore free for general use.
While the advice and information in this book are believed to be true and accurate at the date of
publication, neither the authors nor the editors nor the publisher can accept any legal responsibility for
any errors or omissions that may be made. The publisher makes no warranty, express or implied, with
respect to the material contained herein.
Printed on acid-free paper
Springer is part of Springer Science+Business Media (www.springer.com)
Preface
The content of this volume is intended to place plants within the context of their
surrounding environment, including both biotic and abiotic interactions. Interactions between plants and their environment occur across multiple scales in space
and time, and as the Editor of the volume, I strived to invite and assemble a series of
chapters that cover interactive scales from the organism to the ecosystem and that
are driven by processes spanning seconds to decades. Understanding the fact that
plantenvironment interactions span multiple spatiotemporal scales and that the
processes that control these interactions change with scale is a useful point of
departure for deeper investigations within the field of ecology. This understanding
lies at the foundation of advanced topics such as plantenvironment feedbacks,
nonlinear responses of plants to climate change, extinction dynamics of plants in
fragmented landscapes, and earth system modeling. Starting from this point of
understanding, we can develop strategies for effective management and conservation of natural resources in the face of the daunting environmental challenges that
we face as a global society. The continuity of topics from fundamental ecology to
sustainable protection of ecosystems is crucial as a theme and pedagogic framework in the academic courses offered to undergraduate students in the plant
sciences. Nearly all topics involving plant ecology can be developed within the
conceptual framework of spatiotemporal scaling. This book has been prepared with
this conceptual framework in mind. In all chapters, we have tried to make connections from smaller to larger scales of ecological organization. We tried to communicate the fundamental nature of these connections in as simple and clear a manner
as was possible as a means to reach mid-program to advanced-program undergraduate students, the primary intended audiences for this book.
The book is divided informally into three sections. In the first eight chapters,
fundamental principles of plantenvironment interactions are discussed.
In the first chapter, Reichstein et al. provide an overview of the scales and types
of interactions that determine how plants respond to their environments. Topics in
this chapter extend from global productivity to organismic phenology. A common
theme is control over organism and ecosystem dynamics by climate, and an
emphasis is placed on integrating observations with computer modeling as a
means of understanding ecological processes across multiple scales.
The chapter by Bierzychudek takes up the topic of plant populations and the
factors that control their persistence. Important factors discussed in this chapter
v
vi
Preface
Preface
vii
in climate that threaten forest sustenance through increased frequencies of largescale insect attacks, increased numbers and sizes of wildfires, and exploitation of
wood and water resources.
Sandquist takes up the topic of plants in desert ecosystems. He develops the
concept that plants have evolved highly unique adaptive strategies to deal with the
extremes of heat and drought in desert climates. The novel nature of desert plant
adaptations has fueled the curiosity of plant ecologists for the last two millennia and
provides clear examples of how form and function must be considered together as
the adaptive clay that is sculpted by natural selection.
The chapter by Germino takes us to another extreme of environmental tolerance
that of the short growing seasons and cold temperatures in alpine ecosystems.
Plants in these ecosystems have evolved unique morphological forms that allow
them to persist in the warmer surface boundary layer next to the ground and thus
become uncoupled from the cold temperatures that occur higher up. In both deserts
and alpine ecosystems, seedling establishment is difficult and infrequent, and so
disturbance due to biotic and abiotic stresses have the potential to exert long-term
impacts on community composition and ecosystem processes.
The chapter by Peterson takes us to another example of abiotic extremes in
discussing the ecology of arctic ecosystems. In these high-latitude regions, cold
temperatures slow the rate of decomposition and create extremely low levels of soil
fertility. Animals take on novel facilitative roles that redistribute and recycle
nutrients, and unique plant adaptations have evolved to provide access to nutrient
sources that are not commonly used in temperate ecosystems.
Blair et al. discuss the nature of grasslands. Grassland communities have high
root-to-shoot ratios and are maintained by climate, fire, and frequent disturbance
due to grazing. Together, these processes provide natural impediments to the
invasion of woody species. However, when these natural mechanisms break
down due to overgrazing or landscape fragmentation from human land use, community dynamics can shift, allowing invasion of both woody and nonwoody exotic
species. This chapter on grassland ecology provides a nice case study on the
challenges we face due to species invasions into novel niches.
Moving to the boundary between terrestrial and ocean ecosystems, Armitage
considers the nature of coastal wetlands and in particular salt marshes and mangrove swamps. As in the case for desert, alpine, and arctic ecosystems, the saline
extremes of these coastal wetlands has produced a type of vegetation with unique
adaptations in this case, adaptations to avoid or tolerate salt uptake. These
ecosystems are extremely vulnerable to the deposition of pollution from human
industries. The direct and indirect effects of this pollution create imbalances in the
availability of oxygen and nutrients, which in turn reduce plant productivity and
threaten food webs.
Kirkman discusses the nature of immersed seagrass ecosystems, moving our
perspective even further offshore. Seagrass communities are among the most
valuable on earth for providing goods and services valued by humans they
represent the natural hatcheries for our most valued seafood fishes. Though the
term seagrass would suggest ecosystems based on a monotypic life form, here we
viii
Preface
find some of the most biologically diverse communities on earth. The same
pollution that threatens near coastal wetlands and swamps, however, has caused
an unraveling of natural species interactions in seagrass ecosystems and has
destabilized the hidden mechanisms that sustain diversity and community structure.
Finally, Geider et al. take us to the deeper ocean biomes, where phytoplankton
ecology emerges as the primary topic of plantenvironment interactions. In a rather
comprehensive treatment, these authors provide details of how marine algae tolerate the near-surface ocean environment characterized by high solar radiation and
low nutrient availability, how oceanographers study these interactions, and how
excess nutrient burdens, climate change, and increases in acidity are capable of
changing ocean productivity and altering the global carbon cycle.
In the final four chapters of the book, we consider some of the issues associated
with plants and their role in environmental sustainability.
Leakey tackles the issue of recent increases in the mean global atmospheric CO2
concentration and its influence on plant photosynthesis and the efficiency by which
water is used. He discusses this topic from the foundations of photosynthetic
biochemistry and stomatal function and describes how environmental changes in
the atmospheric CO2 concentration interact with these processes to influence crop
yield and food security.
Wiedinmyer et al. provide a chapter on plant volatile organic compound emissions and their influences on air quality. In particular, they consider recent increases
in the production of tropospheric ozone and atmospheric aerosols, both of which
affect global climate. It has been known for several decades that the emission of
volatile organic compounds from forests can affect a vast number of atmospheric
chemical reactions. However, the final products of these reactions, such as ozone
and aerosols, have been difficult to quantify primarily because the chemistry has
been studied in theoretical terms. We are just now beginning to accumulate the
results from field campaigns and studies of forests such that accurate quantitative
predictions are becoming possible. This issue is also relevant to our expanded
reliance on global agriforests for wood, pulp, and energy production. Most
agriforest tree species emit relatively high amounts of reactive volatile organic
compounds and are thus capable of affecting regional and global air quality.
OKeefe et al. discuss the development of cellulosic biofuels as an alternative
to our reliance on fossil fuels. Consideration of biofuels within the context of
environmental impacts must be generated from knowledge of total resource use
and the potential for hidden resource costs. These authors take on the complexities
of this issue and consider the costs of biofuel production in comprehensive terms
including the costs of water, nutrients, and overall energy.
In the final chapter of the book, Hamilton provides a new framework for
sustainability science. He focuses specifically on the need for integration of knowledge on natural systems such as that provided in the preceding chapters into the
social, economic, and political discussions that ultimately determine how we
manage our natural resources. His chapter brings us to the conclusion that
human well-being is intricately tied to the relations between societies and natural
Preface
ix
ecosystems and that this nexus, with human well-being as a central concern, should
be the focus of strategies for action that improve natural resource management.
As a member of the baby-boom generation, I have observed immense changes
in the earth system over the past five decades. The population of the earth has nearly
doubled since the year of my birth. From hindsight, it is clear that as the population
of the earth has expanded, the margin for error in how we manage our natural and
agricultural ecosystems has contracted. As future generations take on the responsibility for managing our natural resources, one of the most effective things we can
contribute is our accumulated knowledge organized in a way that educates them
and allows them to avoid some of the catastrophic mistakes that prior generations
have made. This book hopefully provides some movement in that direction.
Although a tendency often exists to attack a problem at the scale of its impact,
knowledge of the processes and interactions that lie beneath the scale of impact will
often lead to better-informed solutions from the bottom-up. Hopefully, the
emphasis on processes and interactions that cross all scales of plantenvironment
interaction, which we have tried to produce in this book, will contribute to future
solutions.
Tucson, AZ, USA
June 2014
Russell K. Monson
Series Preface
xii
Series Preface
It is expected that new editions of the printed version will be published every
35 years. The project is proceeding volume by volume, with volumes appearing as
they are completed. This also helps to keep the text fresher and the project more
dynamic.
We would like to thank our host institutions, colleagues, students, and funding
agencies, who have all helped us in various ways and thus facilitated the development of this series. We hope this volume is used widely and look forward to seeing
it develop further in the coming years.
King Abdullah University of Science & Technology,
Thuwal, Saudi Arabia
School of Plant Sciences, University of Arizona,
Tucson, AZ, USA
22 July 2014
Mark Tester
Richard Jorgensen
Editor Biography
xiii
Mark Tester is Professor of Bioscience in the Center for Desert Agriculture and the
Division of Biological and Environmental Sciences and Engineering, King Abdullah
University for Science and Technology (KAUST), Saudi Arabia. He was
previously in Adelaide, where he was a Research Professor in the Australian Centre
for Plant Functional Genomics and Director of the Australian Plant Phenomics
Facility. Mark led the establishment of this Facility, a $55 m organisation that
develops and delivers state-of-the-art phenotyping facilities, including The Plant
Accelerator, an innovative plant growth and analysis facility. In Australia, he led a
research group in which forward and reverse genetic approaches were used to understand salinity tolerance and how to improve this in crops such as wheat and barley.
He moved to KAUST in February 2013, where this work is continuing, expanding also
into work on the salinity tolerance of tomatoes.
Mark Tester has established a research program with the aim of elucidating the
molecular mechanisms that enable certain plants to thrive in sub-optimal soil
conditions, in particular in soils with high salinity. The ultimate applied aim is to
modify crop plants in order to increase productivity on such soils, with consequent
improvement of yield in both developed and developing countries. The ultimate
intellectual aim is to understand the control and co-ordination of whole plant
xv
xvi
xvii
Contents
29
67
...........................
89
119
Plant-Microbe Interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
David A. Lipson and Scott T. Kelley
143
177
205
247
10
............................
273
11
Plants in Deserts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Darren R. Sandquist
297
12
327
xix
xx
Contents
13
363
14
Grassland Ecology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
John Blair, Jesse Nippert, and John Briggs
389
15
425
16
457
17
483
18
533
19
573
20
601
21
631
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
655
Contributors
xxii
Contributors
Contributors
xxiii
Markus Reichstein Department of Biogeochemical Integration, Max-PlanckInstitute for Biogeochemistry, Jena, Germany
Andrew D. Richardson Department of Organismic and Evolutionary Biology,
Harvard University, Cambridge, MA, USA
Darren R. Sandquist Department of Biological Science, California State University, Fullerton, CA, USA
Clint J. Springer Department of Biology, Saint Josephs University, Philadelphia,
PA, USA
Allison Steiner Department of Atmospheric, Oceanic and Space Sciences,
University of Michigan, Ann Arbor, MI, USA
David J. Suggett Functional Plant Biology & Climate Change Cluster, University
of Technology, Sydney, NSW, Australia
Amy M. Trowbridge Department of Biology, Indiana University, Bloomington,
IN, USA
Christine Wiedinmyer Atmospheric Chemistry Division, NCAR Earth System
Laboratory, National Center for Atmospheric Research, Boulder, CO, USA
Contents
Environmental Controls on Vegetation: Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Environmental Controls: Climate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Environmental Controls: CO2, O3, Pollutants, and Nitrogen Deposition . . . . . . . . . . . . . . . . . . .
Ozone and Air Pollutants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Soil Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Animals Including Humans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Plant Responses to the Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Influences of Vegetation on Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Microclimate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Transpiration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Surface Energy Budget . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3
5
7
9
9
10
11
12
12
13
14
M. Reichstein (*)
Department of Biogeochemical Integration, Max-Planck-Institute for Biogeochemistry, Jena,
Germany
e-mail: mreichstein@bgc-jena.mpg.de
A.D. Richardson
Department of Organismic and Evolutionary Biology, Harvard University, Cambridge, MA, USA
e-mail: arichardson@oeb.harvard.edu
M. Migliavacca
Department of Biogeochemical Integration, Max-Planck-Institute for Biogeochemistry, Jena,
Germany
Department of Earth and Environmental Science, University of Milano-Bicocca, Milan, Italy
e-mail: mmiglia@bgc-jena.mpg.de
N. Carvalhais
Department of Biogeochemical Integration, Max-Planck-Institute for Biogeochemistry, Jena,
Germany
Departamento de Ciencias e Engenharia do Ambiente, DCEA, Faculdade de Ciencias e
Tecnologia, FCT, Universidade Nova de Lisboa, Caparica, Portugal
e-mail: ncarval@bgc-jena.mpg.de
# Springer Science+Business Media New York 2014
R.K. Monson (ed.), Ecology and the Environment, The Plant Sciences 8,
DOI 10.1007/978-1-4614-7501-9_22
M. Reichstein et al.
15
17
17
18
19
19
21
21
22
23
24
25
25
Abstract
It has been known for a long time that the environment shapes the appearance
of vegetation (vegetation structure). The systematic description of these effects
has led to classifications of life forms at the organismic scale and biomes at the
global scale by Alexander von Humboldt, Christen C. Raunkir, Wladimir
Koppen, and other early plant geographers and plant ecologists.
Consequently, plant traits and processes carried out by plants (vegetation
function) are influenced by climate and other environmental conditions.
However, given the previous limitations of both observations and theory,
systematic and comparative studies of plant ecology and physiological ecology only began in the twentieth century.
Through their adaptive and genetic constitutions, plants can react to
environmental changes by different mechanisms involving various time
scales. These mechanisms include acclimation, plasticity, and evolution.
Plant reactions, in turn, can feed back to influence the environment at
different scales by exchanges of matter and energy. For example, plants
humidify the air, change turbulence and wind field, and hence influence
cloud formation; they absorb carbon dioxide, produce oxygen and reactive
volatile organic compounds, and modify, protect, and stabilize soils.
There are a large variety of techniques available to researchers for the
observation of vegetationenvironment interactions at different time scales.
No single technique can answer all questions; they have to be used synergistically, and often times these suites of observations have to be deployed
across broad geographic areas and in multiple types of biomes.
Due to the complexity of interactions and feedbacks between vegetation and
the environment, numerical modeling has become a pivotal tool in conjunction with modeldata fusion techniques. This new emphasis on fusing observations and theory has provided scientists with unprecedented insight into the
mechanisms governing plantatmosphere interactions, permitted the scaling
of mechanisms across broad spans of space and time, and provided an
integrated picture of global ecological processes.
M. Reichstein et al.
Fig. 1 Different global views on similar spatial patterns of climate and vegetation. (a) Climate
classification by Koppen (1923), update by Kottek et al. (2006). (b) Remote sensing view from the
NASA MODIS sensor (From http://svs.gsfc.nasa.gov/vis/a000000/a003100/a003191/frames/2
048x1024/background-bluemarble.png). (c) Annual carbon dioxide uptake by photosynthesis of
vegetation (GPP) inferred from a statistical model, derived from ground observations and remote
sensing
1,000
10
100
LNA (g
10,000
10
ma
0.1
1,000
100
10,000
m 2)
ol
100
1,000
LNA (g m 2
)
10
ma
ss
1
10
10
(nm
(%
)
0.1
1
d
100
10
10,000
10
100
1,000
LNA (g m 2
)
10
10,000
(g
(%
ss
1
100
1,000
LNA (g m 2
)
ma
10
0.001
m 2
)
0.1
1
0.1
are
0.01
10
0.1
Pmass (%)
10
g 1 ss
s 1
)
LL (months)
100
100
1,000
Fig. 2 Three-way trait relationships among the six leaf traits with reference to LMA, one of the
key traits in the leaf economics spectrum. The direction of the data cloud in three-dimensional
space can be ascertained from the shadows projected on the floor and walls of the threedimensional space. LMA leaf mass per area, P phosphorus, N nitrogen, Amax light-saturated
photosynthesis, R respiration, LL leaf life span (From Wright et al. 2004)
M. Reichstein et al.
depends on the leaf area index (LAI) of the ecosystem (defined as the amount of
one-sided green leaf area per unit ground surface area, m2/m2) and on the architecture of the canopy, and it is converted into chemical energy in sugars and secondary
metabolites. Photosynthetic processes are affected not only by the amount of PAR
but also by its quality. Recent studies showed higher ecosystem CO2 assimilation
efficiency under skylight conditions that foster a high fraction of diffuse radiation
(Mercado et al. 2009). A more uniform distribution of irradiance causes an increase
in the proportion of light penetration through the canopy and irradiance per unit of
LAI, once again illustrating the interaction between a driving environmental variable, vegetation (or, in this case, canopy) structure, and a physiological variable,
such as CO2 assimilation rate. Moreover, at the canopy level the redistribution of the
solar radiation load from photosynthetically light-saturated leaves to non-saturated
(or shaded) leaves results in a greater increase in leaf photosynthesis rate. This is due
to the fact that shaded leaves conduct most of their photosynthetic CO2 assimilation
in the interactive domain located in the linear part of the light curve response
(approximating a first-order relationship with absorbed radiant energy), while the
saturated, sunlit leaves operate in the interactive domain located in the plateau of
the light response curve (approximating a zero-order relationship with absorbed
radiant energy). SR directly/indirectly influences many secondary plant processes
such as seedling regeneration, leaf morphology, and the vertical structure of stands.
The seasonal variation of photoperiod is also an important factor controlling both
leaf flush and leaf senescence and therefore, together with temperature and water
availability, controls plant phenology and the growing season length.
From the molecular to ecosystem scales, temperature influences biological
processes by controlling the kinetics of enzyme-catalyzed chemical reactions and
thus controlling the rates of plant growth, the patterns of seasonal phenology in
ecosystems, the distribution of species and diversity of communities, and the
decomposition and mineralization of soil organic matter. Generally, the control by
temperature causes process kinetics to exhibit an optimum at intermediate temperatures. The response of processes to temperature variations can be flexible, leading
to time-dependent acclimation responses that allow for maintaining the performance
of processes across a range of temperature conditions (Atkin et al. 2005).
Aside from direct impacts on ecosystems, increasing temperatures can trigger
indirect effects on plants in the ecosystem; many of which interact with one another
to produce subtle synergies. On the one hand, warmer temperatures may enhance
decomposition, releasing nutrients through mineralization; on the other hand,
enhanced evaporation may decrease soil water content, reducing decomposition
rates and its consequent release of nutrients and decreasing the mobility of nutrients
from the soil into plants. As another example, on the one hand, warmer springs, as a
result of climate change, can induce plants in temperate-latitude biomes to initiate
their seasonal growth earlier and thus increase their potential to assimilate CO2
from the atmosphere; but warmer autumns can also potentially interfere with coldtemperature hardening, placing plants at increased risk of physiological damage
during a critical phase of seasonality when frosts are interspersed with favorable
weather.
Precipitation is another of the crucial environmental drivers of ecosystem functioning at different spatial and temporal scales. At short time scales, precipitation and
soil water content control stomatal conductance, and because stomatal conductance
is coupled with photosynthesis, soil water thus influences the rate of CO2 assimilation by vegetation. At longer time scales, the depletion of soil water content due to
scarce precipitation may lead to prolonged water stress with a consequent modification of vegetation structure, such as leaf area index, rooting depth, and chlorophyll
content. Since higher plants do not directly rely on precipitation but rather on water
stored in the soil, the timing of precipitation in relation to the evaporative demand of
the atmosphere, and thus mean air temperature, is of high importance.
Relative humidity (rH) is defined as the ratio of actual water vapor content to the
saturated water vapor content at a given temperature and pressure. rH determines the
vapor pressure deficit (VPD) between the soil and atmosphere and between the plant
and atmosphere, and thus, climate and the spatial distribution of humidity in the
atmosphere control potential evaporation rates and surface energy budgets at the
global scale. The VPD directly influences plant water relations and indirectly affects
hydraulic connectivity between leaves and the soil, leaf growth, photosynthesis, and
evapotranspiration processes through stomatal control and leaf water potential.
Wind speed is another key factor controlling vegetation processes. Different
regimes of wind speed and direction may influence physiological and mechanical
aspects of vegetation. The main physiological effects are related to an enhancement
of evapotranspiration. Wind removes the more humid air around the leaf by
replacing it with drier air and, thus, increases the rate of transpiration. Finally,
wind speed influences photosynthesis rates. Turbulence increases with wind speed
in the atmosphere, which mixes CO2 from higher levels in the atmosphere downward toward the canopy, and thus increases the availability of CO2 for photosynthesis. Turbulence also mixes heat energy between the canopy surface and areas
higher in the atmosphere, affecting the potential for vegetated surfaces to exchange
sensible heat (through convection) with the atmosphere and thus contribute to the
surface energy balance. Wind may also have mechanical impacts on vegetation by
damaging shoots, controlling the allocation of carbon to stem thickening, and
controlling the timing and patterns of leaf, flower, and fruit shedding. Crops and
trees with shallow roots may be uprooted, leading to other secondary effects such as
soil erosion, nutrient deposition, and recruitment opportunities for seedlings requiring a gap in the vegetated canopy. At the landscape scale, high wind speeds,
associated with conditions of low rH and moisture of vegetation, may also contribute to vegetation drying and thus enhancement of the ignition potential of wildfires
and, once ignited, the spread and intensity of fires.
M. Reichstein et al.
Since the CO2 concentration in the intercellular air spaces of the leaf is about
70 % of atmospheric CO2, leaf photosynthesis is expected to respond positively
to the atmospheric increase of CO2 observed since the preindustrial era, which is
related to the increase of anthropogenic emissions from fossil fuel combustion
and land-use change. Empirical evidence from CO2 fumigation experiments
(FACE, Free-Air CO2 Enrichment studies) has shown that the expected increase
of CO2 concentration in the atmosphere of the future enhances plant growth, the
so-called CO2 fertilization effect (Norby and Zak 2011). These studies have
also revealed a response of leaf photosynthesis to elevated CO2 that is dependent
on the conditions at which the plant was grown. In essence, plants grown at
elevated CO2 accumulate sugars at a greater rate than those grown at lower
atmospheric CO2 concentrations. The accumulated sugars trigger changes in the
expression of the genes for Rubisco, the primary CO2-fixing enzyme of photosynthesis, such that fewer enzyme molecules are produced. Rubisco is the most
abundant protein on Earth, and its production by plants utilizes approximately
30 % of the nitrogen resource available to plants. At elevated CO2, a reduction in
the allocation of nitrogen to the production of Rubisco per unit of leaf area
means that more nitrogen can be allocated to the production of new leaf area.
Thus, the high-CO2 feedback enhances the nitrogen-use efficiency of plants
and enhances the potential growth rate of plants in an elevated CO2 (future)
atmosphere.
Besides the increase of CO2, anthropogenic activities cause an increase in
atmospheric nitrogen (N) deposition, particularly of nitrogen oxide compounds
(NOx), and N input to the biosphere caused by the use of fertilizers. The combustion
of fossil fuels and the burning of biomass associated with forest clearing and
agricultural development tend to create a high-temperature process, called the
Zeldovich reaction, which scrambles the N released from plant tissues with the
O2 consumed from the atmosphere and creates NOx compounds that are deposited
back to ecosystems. Once deposited to the soil, microorganisms can convert the
deposited NOx to nitrate and ammonium ions, capable of plant uptake. Due to their
tendency to be leached from soils, nitrate and ammonium are scarce in natural,
unperturbed ecosystems and play a critical role in the biosphere by determining the
potential rates of primary productivity. N availability especially limits gross primary productivity (GPP) and terrestrial carbon (C) sequestration in the boreal and
temperate zone. Human activities associated with the burning of fossil fuels and the
production of agricultural fertilizers have doubled the input of N since 1860.
These anthropogenic changes have had consequences for the turnover of N and
storage of C. In particular, an enhancement of forest growth associated with N
fertilization and a reduction of soil respiration (Janssens et al. 2010) have been
observed. The terrestrial C and N cycles are tightly related. At low N availability, a
doubled CO2 concentration shows a small effect on biomass and photosynthetic
rates, with a negative feedback due to the sequestration of N into the increment
of biomass: the CO2 fertilization increases the terrestrial C storage, as well as
the terrestrial N stock, with a consequent reduction of N availability in the soil
(Zaehle 2013).
Soil Properties
Soils have a fundamental influence on vegetation by providing the most important
reservoir of nutrients and water needed for the biological activities of plants, as well
as serving as a medium for structural anchorage. Soils are more than the inorganic
products of crushed and weathered rocks; rather soils are living systems, a dynamic
component of the Earth system because of the organisms they hold (Bahn
et al. 2010). Carbon is exuded by roots and root-associated fungi, and these
exudates supply carbon to heterotrophic bacteria and other microorganisms that
in turn mineralize soil organic matter, freeing nutrients to be reabsorbed by plants.
In fact, plants must be considered as part of the soil (through their roots). There are
physical, chemical, and biological soil factors that exert profound influences on
vegetation. The main physical characteristics are soil texture, structure, and depth.
Soil texture is determined by the content of silt, clay, and sand, as well as larger
solid matter such as gravel and rocks. Soil texture determines the water holding
capacity of soils, hydraulic conductivity through soils to roots, and the cation
exchange capacity of a soil. Soil depth is determined by the position of the bedrock
10
M. Reichstein et al.
or of the water table and by site characteristics such as slope and topography. Soil
depth and its association with soil organic matter content also determine the portion
of soil usable by plant roots and, therefore, the total water and nutrient holding
capacities.
Chemical characteristics of soils include fertility and acidity (pH), which influence the capacity of soil to sustain growth and maintenance of metabolism in plants.
Soil pH affects the availability of macro- and micronutrients by controlling the
chemical forms of the nutrients. The optimum soil pH range for most plants is
between 5.5 and 7.0, although many plants have adapted to pH values outside this
range. The concentration of available N is less sensitive to pH than the concentration of available phosphorus (P). In order for P to be available for plants, soil pH
needs to be in the range 6.07.5. If pH is lower than 6, P starts forming insoluble
compounds with iron and aluminum, and if pH is higher than 7.5, P starts forming
insoluble compounds with calcium.
11
12
M. Reichstein et al.
Microclimate
Plants influence microclimate in numerous ways. Forest trees provide perhaps the best
example, because their vertical trunks and elevated foliage create unique threedimensional gradients of environmental conditions and resource availability from the
top of the canopy to the forest floor. The evolution of tall, woody plants was therefore a
critical event for life on our planet because it resulted in remarkable habitat diversity
through vertical stratification. It created a novel niche within which organisms could
evolve and adapt the vertical niche. Today, this diversity is best exemplified by the
tropical rainforests, in which highly specialized communities of plants and animals are
adapted to different canopy strata, each of which has its own microclimate.
13
Transpiration
Gas exchange for photosynthesis occurs through stomata on the leaf surface.
Stomata open during the day, allowing CO2 to diffuse into the leaf. At the same
14
M. Reichstein et al.
Fig. 3 The potential effect of vegetation on local climate. See text for more information (From
Lyons 2002)
time, however, diffusion of water vapor from inside the leaf to the free atmosphere
is driven by the non-saturated moisture state of the atmosphere and its evaporative
demand for water. Through this process, called transpiration, plants are responsible
for the movement, each day, of massive quantities of water from the soil column to
the atmosphere. Transpiration has a significant cooling effect on surface climate,
removing heat through the process of latent heat exchange. Additionally, at regional
scales, transpiration by plants results in the increased abundance of clouds, which
both moderate surface temperature and enhance precipitation. A nice example of
such effects is the so-called bunny fence experiment in Australia (Fig. 3). The
fence has led to a sharp boundary of vegetation types across a relatively homogenous terrain, with a visible influence on cloud cover.
15
modest increase in albedo (cooling effect) more than offset by a large decrease in
transpiration (warming effect). For a given amount of net radiation, lower latent
heat flux must be offset by higher sensible heat flux. Thus, boreal conifer forests,
which have lower rates of evapotranspiration than boreal deciduous forests, have
higher rates of sensible heat flux, which returns energy to the atmosphere and
promotes the development of a deeper atmospheric boundary layer. When available
soil moisture is reduced during drought, driving reductions in evapotranspiration,
there is similarly a corresponding increase in sensible heat flux, ultimately affecting
mesoscale circulation and atmospheric transport.
16
M. Reichstein et al.
17
well as soybeans and other legumes) and bacteria such as Rhizobium and Frankia
are important because the bacteria can fix N2 gas (which cannot be otherwise used
by plants) from the atmosphere to NH3, or ammonium (NH4+), which is available
for plant uptake. Additionally, plant root exudates can enhance the chemical
weathering of soil minerals to forms that are phyto-available.
Observation Strategies
As discussed above, plantenvironment relations are manifold and operate at
different scales. Accordingly, observation strategies, which span across several
spatial and temporal scales, have been developed and are needed for a complete
exploration of plantenvironment interactions (Fig. 4).
18
M. Reichstein et al.
Fig. 4 Observational systems related to vegetation and ecosystem function across temporal and
spatial scales with special emphasis on carbon balance and trace gas exchange with the atmosphere
19
LAI, aboveground biomass) have been collected in a harmonized way in the context
of a research initiative called TRY, which focuses on the collection of data and
knowledge on plant traits at the global scale (Kattge et al. 2011).
Flux Measurements
Flux measurements, i.e., the measurement of gas exchange between plants and the
surrounding air, can be done at organ, whole-plant, or ecosystem level. They are
classically performed with enclosures (cuvettes or chambers) that surround leaves,
branches, or the whole plant and where air is blown through, and the concentration
differences between the inlet and outlet are measured. These measurements are
very precise, but the presence of the chambers can change the microenvironment
around the object to be measured, e.g., by changing the radiation balance, thus
altering the respective gas exchange. This problem is overcome by eddy covariance
(EC), a micrometeorological technique that relies on the combination of highfrequency measurement (1020 Hz), temperature, wind speed, and gas concentration (e.g., CO2, water vapor, methane, etc.) (Baldocchi 2008). In the last three
decades, this technique has been widely used for monitoring carbon, water, and
energy fluxes and, more recently, fluxes for methane and other greenhouse gases, in
more than 500 research sites, scattered across a variety of biomes and climatic
regions. The long-term measurements of CO2 and greenhouse gas fluxes obtained
using the eddy covariance technique make it a useful tool for elucidating the carbon
balance of terrestrial ecosystems and the causes of its interannual variability and for
improving the understanding of the interaction between carbon, water, energy
fluxes, and climate. Measuring the abundance and fluxes of stable isotopes has
become possible with high temporal resolution and yields complementary information on plant ecophysiology (Griffis 2013).
Remote Sensing
Remote sensing (RS) observations can provide spatial and temporal variability of
ecosystem properties driving carbon, water, and energy fluxes, as well as important
information about vegetation and ecosystem structure (e.g., aboveground biomass,
leaf area index). RS data provides spatial (global, regional, and local) and temporal
(decadal, seasonal, and interannual) information about the important properties of
the ecosystem. Moreover, by using multitemporal classification methods, RS can be
used to gather information about land-use change and disturbance (in particular
fires and deforestation). However, RS data can be hampered by the contamination
of the signal by aerosols and clouds and by the fact that the parameters are estimated
by using empirical relationships or radiative transfer models and not by direct
measurement; whenever models must be inserted into a diagnostic or prognostic
process, gaps in knowledge produce uncertainty in calculation. Nevertheless,
interactions between climate and vegetation type can often be clearly inferred,
20
M. Reichstein et al.
Fig. 5 Remotely sensed images of vegetation cover (top) and land surface temperature (bottom)
before (left) and during (right) the 2003 European heat wave. Denser vegetation cover with forest
yields less surface heating (From Zaitchik et al. 2006)
as in Fig. 5, where strong gradients of land surface temperature are found depending
on vegetation type and density as a consequence of an extreme heat wave.
The typology of measurements and parameters retrieved via RS depends on the
characteristics of the sensors. With the development of hyperspectral imaging
or reflectance sensors, it is possible to look at objects (target) using a vast portion
of the electromagnetic spectrum. Targets such as leaves or tree canopies have
unique fingerprints (spectral signatures) across the electromagnetic spectrum.
By exploiting this information, it is possible to derive important properties such
as chlorophyll/pigments, leaf nitrogen, extractable water content, etc. RS data can
be collected at different spatial scales by using satellite products and airborne
platforms with hyperspectral sensors, as well as in the proximity of the surface
(proximal sensing). Proximal sensing is increasingly growing because it is one way
to better understand the relationships between RS data and ecosystem processes at
high temporal resolution, if associated with EC measurements. An emergent branch
of RS is the direct inference of physiological processes, in particular photosynthesis. Among these, the measurement of sun-induced chlorophyll fluorescence (SIF)
by passive (i.e., without artificial excitation sources) RS systems at field scale has
been proven to be a valuable method for the assessment of plant photosynthesis
(Meroni et al. 2009).
Another important technique, which yields three-dimensional structural information (e.g., leaf and branch distribution), is terrestrial LiDAR scanners (Levick
and Rogers 2008). Terrestrial LiDAR measurements are generally collected using
an instrument placed on a survey tripod above the ground in the experimental site.
Their usage for estimating leaf area stems from a very high spatial resolution and a
relatively small laser footprint size with respect to the typical dimensions of leaves
21
and other tree organs. New LiDAR missions will be used in the future to precisely
describe the canopy structure, in particular the vertical distribution of elements in a
canopy, the tree height, and the tree cover.
Modeling Strategies
From a theoretical point of view, models can be defined as representations of
systems or processes underlying a wide variety of observed properties and functions. Ecosystem models are simplifications of the complex organizational structures and interactions observed in nature but ultimately synthesize our knowledge
and theory. In this regard, very different modeling approaches have been developed, depending on the characteristics or dynamics to be represented or hypotheses
to be explored. These range from simple empirical univariate approaches describing
processes of decomposition or primary productivity to process-based approaches
with a more mechanistic representation of physical chemical reactions in living
organisms to describing dynamics of vegetation changes. But independent of the
22
M. Reichstein et al.
23
24
M. Reichstein et al.
states between observations and model simulations allows inferring the likelihood
of the underlying mechanisms and hypothesis about ecosystem functioning
(Reichstein and Beer 2008). MDF approaches enable the explicit treatment of the
main sources of uncertainty arising from model structure, parameters, initial
conditions, and observational data used in driving or constraining the model (Liu
and Gupta 2007). Model parameters control the sensitivities of ecosystem responses
to environmental conditions but also regulate internal dynamics related, for
instance, to the maximum photosynthetic capacity, optimum temperature for photosynthesis, allocation of carbon to plant organs, surface to leaf area, etc. Although
some of these parameters can be, and have been, measured, there are uncertainties
related to observational methods as well as to its spatial and temporal representativeness, many times translated in the high variability of observations. The model
structure is tested by exploring the likelihood of the model given the observations
within the feasible distributions of parameters. The observational uncertainty can
also be formally integrated in MDF approaches by weighing higher (lower) the
observational records with lower (higher) uncertainties in the cost function. But the
evaluation of the model is very dependent on the construction of the cost function,
and modeling exercises have emphasized the challenges in the comprehensive
representation of ecosystems. Given a multivariate comparison of model outputs
with observations that translate different components of an ecosystem, the construction of an unbiased and comprehensive estimator of likelihood becomes a
challenge per se. If integrating the multivariate observations of carbon and water
fluxes and pools in ecosystem modeling provides a comprehensive test to model
structures, it may also bias parameterizations when the datasets dimensions can
vary orders of magnitude, which would tend to favor model behavior for the most
observed variable(s). Another aspect relates to inconsistencies between datasets,
which could lead to parameterization biases and erroneous identification of poor
model structures. The advantage of MDF lies in its ability to formally account for
all these sources of uncertainties in bringing the theory embedded in models and
observations together (Williams et al. 2009).
Overall, exploring model and data integration approaches reflects the possibility
to test theories and hypotheses about ecosystem functioning corroborated by observations. Given the complexity of ecosystems, a comprehensive analysis values the
overall coherence of our understanding of ecosystem functioning, but that does not
detract from using simpler approaches that target exploring conceptual hypotheses.
Ultimately, the association between ecosystem properties and functional behavior
reflects the potential to extrapolate and scale the representation of ecosystem
functioning.
25
unobserved regions? The link between plant structural types and the seasonality of
phenology has motivated the classification of vegetation according to plant functional types (PFT). This classification assumes similar behavior in responses to
environmental conditions, effects on ecosystem structure, and inherent processes.
But the existing diversity holds a multiplicity of structural and functional characteristics that is well beyond the extent of a classification scheme. The possibility to
move beyond classification schemes relies on the ability to link functional
responses of plants and ecosystems to ubiquitous observations of relevant biotic
and abiotic properties or states (Kattge et al. 2011).
Future Directions
It is evident that plants react to the environment and influence the environment at
different scales, from local to global scale. Direct responses to normal variation are
relatively well understood, but in the future the regional feedbacks between plants
and weather, i.e., the regional coupling between vegetation and the atmosphere,
need to be understood better. These feedbacks are largely mediated through the
water and energy cycles. For example, forest and grasslands were shown to exhibit
very different energy fluxes to the atmosphere during heat waves and drought
(Teuling et al. 2010). This way, they contribute differently to the development
and stabilization of heat waves (Seneviratne et al. 2010). Moreover, direct and
indirect responses of vegetation to extreme conditions need further study, with the
main question, under which conditions irreversible processes like mortality
are triggered? In this context it has recently been argued that vegetation responses
to climate extremes can cause a positive global climate feedback by reducing
the photosynthetic uptake (Reichstein et al. 2013). Last but not least, the fate
of vegetation under a rapidly changing climate, as is being experienced now,
will depend on its ability and velocity to adapt to those changing conditions. This
is currently completely ignored in climate models (Stocker et al. 2013). Thus,
joint studies in genetics, developmental biology, biogeochemistry, and biosphere
modeling (e.g., Scheiter et al. 2013) need to be integrated in future research efforts.
References
Ainsworth EA, Yendrek CR, Sitch S, Collins WJ, Emberson LD. The effects of tropospheric ozone
on net primary production and implications for climate change. Annu Rev Plant Biol.
2012;63:63761.
Amthor JS. The McCreede WitPenning de VriesThornley respiration paradigms: 30 years
later. Ann Bot. 2000;86(1):120.
Atkin OK, Bruhn D, Hurry VM, Tjoelker MG. Evans Review No. 2: The hot and the cold:
unravelling the variable response of plant respiration to temperature. Funct Plant Biol.
2005;32(2):87105.
Bahn M, Janssens IA, Reichstein M, Smith P, Trumbore SE. Soil respiration across scales: towards
an integration of patterns and processes. New Phytol. 2010;186(2):2926.
26
M. Reichstein et al.
Baldocchi D. Turner review No. 15. Breathing of the terrestrial biosphere: lessons learned from a
global network of carbon dioxide flux measurement systems. Aust J Bot. 2008;56(1):126.
Barbour MG, Burk JH, Pitts WD, Gilliam FS, Schwartz MS. Terrestrial plant ecology. 3rd
ed. Menlo Park: Addison Wesley Longman; 1999.
Beerling DJ. The emerald planet. New York: Oxford University Press; 2007.
Bonan GB. Ecological climatology concepts and application. 2nd ed. Cambridge, UK: Cambridge University Press; 2008a.
Bonan GB. Forests and climate change: forcings, feedbacks, and the climate benefits of forests.
Science. 2008b;320(5882):14449.
Bowman DMJS, et al. Fire in the earth system. Science. 2009;324(5926):4814.
, Dybzinski
Franklin O, Johansson J, Dewar RC, Dieckmann U, McMurtrie RE, Brannstrom A
R. Modeling carbon allocation in trees: a search for principles. Tree Physiol.
2012;32(6):64866.
Greene EL. Landmarks of botanical history: a study of certain epochs in the development of the
science of botany: part 1, Prior to 1562 A.D. Washington, DC: Smithsonian Institution; 1909.
Griffis TJ. Tracing the flow of carbon dioxide and water vapor between the biosphere and
atmosphere: a review of optical isotope techniques and their application. Agr Forest Meteorol.
2013;174175(0):85109.
Heimann M, Kaminski T. Inverse modeling approaches to infer surface trace gas fluxes from
observed atmospheric mixing ratios. In: Bouwman A-F, editor. Approaches to scaling of trace
gas fluxes in ecosystems. Amsterdam: Elsevier; 1999. p. 27595.
Heimann M, Reichstein M. Terrestrial ecosystem carbon dynamics and climate feedbacks. Nature.
2008;451(7176):4.
Holdridge LR. Determination of world plant formations from simple climatic data. Science.
1947;105(2727):3678.
Janssens IA, et al. Reduction of forest soil respiration in response to nitrogen deposition. Nat
Geosci. 2010;3(5):31522.
Kattge J, et al. TRY a global database of plant traits. Glob Chang Biol. 2011;17(9):290535.
Koppen W. Dle Klirnate der Erde. Berlin: Walter de Gruyt; 1923.
Kottek M, Grieser J, Beck C, Rudolf B, Rubel F. World Map of the Koppen-Geiger climate
classification updated. Meteorol Z. 2006;15(3):25963.
Levick SR, Rogers KH. Structural biodiversity monitoring in savanna ecosystems: integrating
LiDAR and high resolution imagery through object-based image analysis. In: Blaschke T,
Lang S, Hay G, editors. Object-based image analysis. Berlin/Heidelberg: Springer; 2008.
p. 47791.
Liu Y, Gupta HV. Uncertainty in hydrologic modeling: toward an integrated data assimilation
framework. Water Resour Res. 2007;43:W07401.
Lyons T. Clouds prefer native vegetation. Meteorol Atmos Phys. 2002;80(14):13140.
Malmstrom C. Ecologists study the interactions of organisms and their environment. Nat Educ
Knowl. 2010;3(10):88.
McPherson RA. A review of vegetation atmosphere interactions and their influences on
mesoscale phenomena. Prog Phys Geogr. 2007;31(3):26185.
Mercado LM, Bellouin N, Sitch S, Boucher O, Huntingford C, Wild M, Cox PM. Impact of
changes in diffuse radiation on the global land carbon sink. Nature. 2009;458(7241):10147.
Meroni M, Rossini M, Guanter L, Alonso L, Rascher U, Colombo R, Moreno J. Remote sensing of
solar-induced chlorophyll fluorescence: review of methods and applications. Remote Sens
Environ. 2009;113(10):203751.
Monteith JL. Solar-radiation and productivity in tropical ecosystems. J Appl Ecol.
1972;9(3):74766.
Norby RJ, Zak DR. Ecological lessons from free-air CO2 enrichment (FACE) experiments. Annu
Rev Ecol Evol Syst. 2011;42(1):181203.
27
Penuelas J, Staudt M. BVOCs and global change. Trends Plant Sci. 2010;15(3):13344.
Pielke Sr RA, Avissar R, Raupach M, Dolman AJ, Zeng X, Denning AS. Interactions between the
atmosphere and terrestrial ecosystems: influence on weather and climate. Glob Chang Biol.
1998;4(5):46175.
Prentice IC, Bondeau A, Cramer W, Harrison SP, Hickler T, Lucht W, Sitch S, Smith B, Sykes
MT. Dynamic global vegetation modelling: quantifying terrestrial ecosystem responses to
large-scale environmental change. In: Canadell JG, editor. Terrestrial ecosystems in a changing
world. Berlin/New York: Springer; 2007.
Raunkir C. The life forms of plants and statistical plant geography, being the collected papers of
C. Raunkir. Oxford: Oxford University Press; 1934.
Reichstein M, Beer C. Soil respiration across scales: the importance of a modeldata integration
framework for data interpretation. J Plant Nutr Soil Sci. 2008;171(3):34454.
Reichstein M, Bahn M, Ciais P, Frank D, Mahecha MD, Seneviratne SI, Zscheischler J, Beer C,
Buchmann N, Frank DC. Climate extremes and the carbon cycle. Nature. 2013;500
(7462):28795.
Richardson LF. Weather predication by numerical process. Cambridge University Press; 1922.
236 pp.
Richardson AD, Keenan TF, Migliavacca M, Ryu Y, Sonnentag O, Toomey M. Climate change,
phenology, and phenological control of vegetation feedbacks to the climate system. Agr Forest
Meteorol. 2013;169:15673.
Scheiter S, Langan L, Higgins SI. Next-generation dynamic global vegetation models: learning
from community ecology. New Phytologist. 2013;198(3):95769.
Schulze E-D, Beck E, Muller-Hoenestein K. Plant ecology. New York: Springer; 2005.
Seneviratne SI, Corti T, Davin EL, Hirschi M, Jaeger EB, Lehner I, Orlowsky B, Teuling
AJ. Investigating soil moistureclimate interactions in a changing climate: a review. EarthScience Rev. 2010;99(3):12561.
Stocker TF, Dahe Q, Plattner G-K. Climate change 2013: the physical science basis. Working
Group I Contribution to the Fifth Assessment Report of the Intergovernmental Panel on
Climate Change. Summary for Policymakers. IPCC; 2013.
Teuling AJ, Seneviratne SI, Stockli R, Reichstein M, Moors E, Ciais P, Luyssaert S, van den
Hurk B, Ammann C, Bernhofer C, Dellwik E, Gianelle D, Gielen B, Gr
unwald T, Klumpp K,
Montagnani L, Moureaux C, Sottocornola M, Wohlfahrt G. Contrasting response of European
forest and grassland energy exchange to heatwaves. Nat Geosci. 2010;3:7227.
Williams M, et al. Improving land surface models with FLUXNET data. Biogeosciences.
2009;6(7):134159.
Wright IJ, Reich PB, Westoby M, Ackerly DD, Baruch Z, Bongers F, Cavender-Bares J, Chapin T,
Cornelissen JH, Diemer M. The worldwide leaf economics spectrum. Nature. 2004;428
(6985):8217.
Zaehle S. Terrestrial nitrogencarbon cycle interactions at the global scale. Philos Trans R Soc B
Biol Sci. 2013;368(1621).
Zaitchik BF, Macalady AK, Bonneau LR, Smith RB. Europes 2003 heat wave: a satellite view of
impacts and landatmosphere feedbacks. Int J Climatol. 2006;26(6):74369.
Further Reading
Larcher W. Physiological plant ecology: ecophysiology and stress physiology of functional
groups. Springer; 2003.
Purkis SJ, Klemas VV. Remote sensing and global environmental change. Chichester: Wiley;
2010.
Contents
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Structure of Plant Populations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Temporal Patterns of Population Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Causes of Different Temporal Patterns of Plant Population Dynamics . . . . . . . . . . . . . . . . . . . . . . . .
What Forces Regulate the Sizes of Plant Populations? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
The Role of Stochastic Influences, Especially in Small Populations . . . . . . . . . . . . . . . . . . . . . . .
Incorporating Population Structure into Models and Analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Spatial Patterns of Population Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A Brief Guide to Methodological Approaches Used in Field Studies of
Plant Population Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Defining the Boundaries of a Population . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Censusing Populations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Future Directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
30
32
37
38
42
46
49
59
61
61
61
63
63
Abstract
Population dynamics is the study of how and why population sizes change
over time.
Repeated censuses of individuals within populations are the core data collected by plant ecologists studying population dynamics.
Plant populations are characterized by their size (or density) and their structure (the numbers of individuals of different ages and sizes).
Plant population ecologists use observations, experiments, and mathematical
models to document and understand patterns of population dynamics.
Most plant populations appear to be regulated by density-dependent forces;
resource competition and natural enemies are the most likely forces responsible for regulation.
P. Bierzychudek (*)
Department of Biology, MSC 53, Lewis & Clark College, Portland, OR, USA
e-mail: bierzych@lclark.edu
# Springer Science+Business Media New York 2014
R.K. Monson (ed.), Ecology and the Environment, The Plant Sciences 8,
DOI 10.1007/978-1-4614-7501-9_15
29
30
P. Bierzychudek
Introduction
The Haleakala silversword, Argyroxiphium sandwicense subsp. macrocephalum, is
an unusual plant for many reasons, not the least of which is its striking appearance,
like the offspring of a marriage between a footstool and a pincushion (Fig. 1).
Found only on Mt. Haleakala, a dormant volcanic cinder cone on the Hawaiian
island of Maui, this remarkable plant lives on mostly barren, rocky, unstable slopes
at elevations of 2,1003,000 m. Individuals live for up to 50 years before sending up
a flowering stalk that bears as many as 600 flower heads. After this one reproductive
episode, the plant dies.
The Haleakala silversword population has survived the cattle and goats that once
grazed the mountain and persists despite the fact that tourists impressed by their
bizarre appearance once routinely bowled these plants down the mountainside or
uprooted them for souvenirs. Protection from these threats in the 1930s greatly
increased the silverswords numbers over the next 60 years. By the late 1990s, the
silversword population was estimated to be 16 times larger than it had been in 1935,
and this iconic plant came to be considered one of the Hawaiian Islands conservation success stories. However, since the mid-1990s, the silversword population is
once again in decline (Fig. 2; Krushelnycky et al. 2013).
These trends would not have been apparent except for observers who chose to
census the number of silversword individuals in the Mt. Haleakala population,
starting with park ranger S.H. Lamb in 1935 (U.S. Fish and Wildlife Service 1997).
Census data are key to understanding the dynamics of plant populations, i.e., how
numbers of individuals change over time, and to determining the causes of those
changes. This chapter will examine the history, key concepts, main methodologies,
and important unanswered questions in the field of plant population dynamics.
A population is a group of individuals belonging to the same species, living
in the same area. The study of plant population dynamics, i.e., how and why plant
populations change in numbers over time, is a relatively recent chapter in
plant ecology. While a few earlier workers had carried out repeated censuses of
plant populations, British ecologist John L. Harper (19252009) revolutionized
31
Fig. 1 A flowering
Haleakala silversword (Photo
by Forest and Kim Starr)
Fig. 2 Numbers of Haleakala silversword individuals at a high-elevation canyon rim site (open
squares) and at five sampling areas on the crater floor (other symbols) (Figure from Krushelnycky
et al. 2013)
32
P. Bierzychudek
how ecologists thought about plants with his 1967 paper, A Darwinian Approach
to Plant Ecology, and his 1977 book, Population Biology of Plants. Before Harper,
it was mostly zoologists, not botanists, who studied the biology of populations.
Harper, his students, and many other ecologists he influenced developed the
quantitative, process-oriented, and often experimental approach to the study of
plant population dynamics that characterizes the field today. In fact, John Harper
argued that plants were more suitable than most animals for the study of population
dynamics because plants stand still to be counted and do not have to be trapped,
shot, chased, or estimated (Harper 1977, p. v).
Plant population ecologists are interested in knowing what trends characterize
plant populations over time do they increase? Decrease? Remain constant? Are
these patterns predictable or stochastic? What forces are responsible for the different patterns? These questions are of interest not only for their own sake, but also
because their answers can lead to effective problem solving in the fields of
agriculture, forestry, range management, natural area management, and species
conservation.
This chapter will begin by describing the structure of plant populations and by
considering some aspects of plant biology that affect how plant populations are
studied, such as the relationship between size and age, and how individuals are
defined. This will be followed by a description of some of the spatial and temporal
patterns displayed by different populations and a consideration of the possible
causes of these different patterns. The chapter will briefly review some of the
primary methodological approaches used to study plant populations in the field.
Throughout, it will illustrate some of the ways these approaches have been applied
to address particular practical problems, especially in the area of biodiversity
conservation.
33
independent units are distinct genets. Individuals that are physiologically independent of one another are considered separate ramets, regardless of their genetic
similarity. The number of genets in a plant population can be much lower than the
number of ramets. Ramets are often easier to recognize than genets, so this definition
of an individual is more frequently used. Because the identification of individuals
can be so challenging in many species, studies of plant populations have historically
been biased toward those species in which individuals are relatively easy to define;
we know much less about species with strong propensities toward vegetative spread
than about species that tend to restrict their growth to the vertical dimension.
Once the issue of how to define an individual has been addressed, there are two
ways to express the size of a population. Sometimes a populations size is described
as the number of individuals it contains; other times it is the populations density that
is reported, i.e., the mean number of individuals per unit of area. It is important to
keep in mind that density is an average measure for the entire population and that
individuals can be distributed in space in three different ways. Individuals of a
species are sometimes spaced regularly, such that the mean density of individuals in
a series of sampling plots is greater than the variance in density among plots. Alder
shrubs in the Alaskan tundra are regularly spaced; Chapin et al. (1989) suggested
that regular spacing is most likely to be found in habitats with low species diversity
and intense competition for resources, like desert or tundra. Rarely, individuals are
randomly distributed in space (Hutchings 1997); in this case the mean density of
individuals among plots is similar to the variance. Finally, individuals are most often
found in a clumped distribution (Hutchings 1997), with the variance in the density of
individuals among sample plots being greater than the mean. A clumped distribution
pattern can occur if the underlying physical environment is heterogeneous, with
individuals clustered within the suitable patches and absent from the unsuitable
ones. It can also arise from the fact that many plant species have rather localized
seed dispersal, so that seedlings are often found in close proximity to their parents.
In addition to variation in their spatial distribution, individuals within a population
can vary in such characteristics as their size, their age, or their sex. These so-called
demographic parameters often have important effects on how each individual contributes to a populations dynamics. Because most plant species have perfect flowers,
there is only one sex in most plant populations; all individuals are hermaphrodites.
In such species, sex is not a particularly important demographic characteristic. Sex is
a more important demographic parameter in many animal populations and in those
plant species with separate sexes. In such species, the ratio of male to female
individuals can strongly affect a populations potential for increase.
In animals with determinate growth, age is a very important demographic
parameter. Individual animals often must reach a certain age before achieving
sexual maturity, and an individuals probabilities of dying and of giving birth
(probabilities often referred to as vital rates) are well correlated with its age.
By contrast, consider a seedling Eucalyptus, a 5 m tall Eucalyptus in the forest
understory, and a mature 100 m tall Eucalyptus tree, each of which has very
different probabilities of dying and of reproducing. While it is certain that the
mature tree is older than the seedling, the age of the understory individual is more
34
thousands of seedlings
800
number of trees
P. Bierzychudek
600
400
200
0
1
4
5
6
size class
10
35
Thousands
of seedlings
80
70
Missing
diameter
classes
Narrowleaf cottonwood
Black cottonwood
No. Trees
60
50
40
30
20
10
0
10
20
30
40
50
60
70
Diameter (cm)
80
90
100
110
120
Fig. 4 Numbers of cottonwood trees of different trunk diameter size classes in the Lamar Valley
of Yellowstone National Park, USA (Reprinted from Beschta 2003)
young seedlings and saplings and preventing the establishment of mature trees. With
the recent return of wolves to the valley, elk have become more wary, rarely venturing
out of the forest into the open floodplain habitat (Beschta 2003), allowing seedling
cottonwoods to survive unbrowsed. While this hypothesis for the cottonwood stage
structure in Yellowstone remains controversial (Winnie 2012), it is clear that the
unexpected size structure of this cottonwood population demands an explanation.
In even-aged populations of agricultural or greenhouse plants, other patterns of
size structure are observed, and it becomes possible to examine how these patterns
develop and change over time. Frequency distributions of seedling weights are
typically approximately normal (Fig. 5, top row).
Variation in seedling size exists because seed sizes are rarely uniform, and the
size of a seed has a strong influence on the size of the seedling that emerges from it
(Hutchings 1997). Over time, as seedlings grow, their weight distributions tend to
become increasingly skewed (Fig. 5, middle, bottom rows), especially at higher
densities, for several reasons (Hutchings 1997). First, there is genetic variation for
growth rate among a group of individuals. Second, the timing of a seedlings
emergence relative to that of its closest neighbors can give certain seedlings an
initial growth advantage or disadvantage. Third, the spacing of a growing plants
immediate neighbors determines the amount of resources available to it. For all
these reasons, many individuals may remain small, spindly, and fail to flower or
produce seeds. This effect is most extreme and rapid in high-density populations
(Fig. 5, right-hand column).
36
P. Bierzychudek
50
Low density, 60/m 2
First harvest
40 (2 weeks post-emergence)
30
20
10
40
10 16
28
40
4 10 16 28
40
plant weights in mg
4 10 16
28
40
Second harvest
(6 weeks post-emergence)
30
20
10
40
Final harvest
(maturity)
30
20
10
0.8
1.6
2.4
0.8
1.6
2.4
0.8
1.6
2.4
plant weights in g
Fig. 5 Frequency distributions of plant weights for flax, Linum usitatissimum, sown at three densities.
Y-axis percent of the population in each weight class. Heavy black bar represents the mean plant
weight. Top row: seedlings harvested 2 weeks after emergence; weights in mg. Middle row: 6-weekold plants; weights in mg. Bottom row: mature plants; weights in g (Figure from Harper (1977))
37
Over time, the death of some of the small individuals in a dense population can
allow other individuals to achieve larger size, and it is common to observe the size
structure of such populations shifting over time as shown in Fig. 5. Mortality
resulting from competition simultaneously alters the population density. Thus
density and individual plant weight change in concert. In many populations this
process of self-thinning has been shown to follow a temporal pattern represented
by the relationship
w cNk
38
P. Bierzychudek
Fig. 6 A hypothetical
population with little or no
change in numbers with time
number of individuals
3500
3000
2500
2000
1500
1000
2006
2008
2010
2000
2002
2004
2001
2003
2005
2007
2009
year
number of individuals
3500
3000
2500
2000
1500
Fig. 7 A hypothetical
population in which numbers
of individuals are increasing
with time
1000
2000
2002
2004
2006
2008
2010
2001
2003
2005
2007
2009
year
Over longer time periods, the same population can display both patterns. Often
superimposed on these trends, and also evident in populations with little overall
change, is an unpredictable wobble in numbers of individuals (Fig. 9).
Fig. 8 A hypothetical
population in which numbers
of individuals are decreasing
with time
39
3500
number of individuals
3000
2500
2000
1500
1000
2002
2000
2004
2006
2008
2010
2001
2003
2005
2007
2009
year
number of individuals
3500
3000
2500
2000
1500
Fig. 9 A hypothetical
population with unpredictable
fluctuations but no overall
trend in numbers with time
1000
2000
2002
2004
2006
2008
2010
2001
2003
2005
2007
2009
year
40
3500
number of individuals
Fig. 10 A hypothetical
population with an initial size,
N0, of 1,500 individuals and
an annual growth rate, , of
1.09
P. Bierzychudek
3000
2500
2000
1500
1000
2000
2002
2004
2006
2008
2010
2001
2003
2005
2007
2009
year
where the populations net reproductive rate, i.e., the ratio of Nt+1 to Nt. In
Eq. 3, is a constant; in other words, this model contains the implicit assumption
that the populations net reproductive rate does not change as a function of the
populations size, and is not influenced by changing environmental conditions. This
model can be generalized to longer time periods:
Nt N0 t
A population with > 1 is increasing geometrically (see Fig. 10), one with < 1
is decreasing geometrically (see Fig. 11), and one with 1 is not changing in size.
Because this model is in the form of a difference equation, it is a particularly apt way
to describe a population whose size grows (or shrinks) in spurts that occur once a
year. This is the case, for example, for annual species in which individuals live for
one growing season, produce seeds, and die at the end of that season, their seeds
germinating at the beginning of the next growing season.
It is also possible to express the hypothesis that the population growth rate is a
constant in continuous time, a form that some readers may find more familiar:
dN
rN
5
dt
In this continuous-time model of exponential (i.e., geometric) population
growth, r is a parameter known as the intrinsic or instantaneous rate of increase
Fig. 11 A hypothetical
population with an initial size,
N0, of 3,000 individuals and
an annual growth rate, , of
0.91
41
3500
number of individuals
3000
2500
2000
1500
1000
2000
2002
2004
2006
2008
2010
2001
2003
2005
2007
2009
year
and is defined as the difference between the per capita birth and death rates.
A growing population has an r > 0, while a declining one has an r < 0. This
model produces the same results as those shown in Figs. 10 and 11 except that
the change in population size is continuous rather than stepwise. For more about
the correspondence between the difference-equation and continuous-time forms of
the geometric/exponential growth model, see Begon et al. (1996).
When a populations dynamics fit the pattern of change in numbers over time
shown in Fig. 10, it suggests that necessary resources are superabundant relative to the
resource requirements of individuals in the population. This pattern can be observed
in plant populations that have recently colonized an environment where competitors
and predators are rare and where resources are temporarily superabundant, such as
species occupying a recently abandoned agricultural field, a newly logged forest, or
the site of a recent fire, flood, or other catastrophic disturbance. Many species are
specifically adapted to these habitats and are rarely seen in other circumstances,
surviving from disturbance to disturbance by means of long-lived seed banks.
However, few populations exhibit a pattern of geometric growth for more than a
short time; no population is capable of increasing forever without limit. One
obvious cause of population decline is a directional change in the suitability of
the environment resulting from successional change, e.g., as a meadow is colonized
by shrubs and trees, herb and grass species decrease in abundance. It is more
challenging to understand changes in numbers that occur in environments that are
not undergoing such obvious environmental change.
Populations that experience a positive growth phase at first are often limited
(eventually) by abiotic or biotic factors. Some of these factors act with an intensity
that is independent of the size of the population subject to them; these are often
referred to as density-independent limiting forces. For example, a severe drought
might cause the death of all of the seedlings whose roots failed to reach a particular
soil depth, no matter whether the density of seedlings was relatively high or
P. Bierzychudek
number of individuals/m2
42
60
40
20
0
1988
1990
1992
1994
1996
1998
year
relatively low. A late frost might cause the abortion of all the developing seeds in a
population. Density-independent mortality can periodically reduce the size of a
population; Fig. 12 shows population trends for Linanthus parryae, a desert annual,
in the Mojave Desert of southern California, USA. The years when no adults were
recorded had extremely low rainfall; the population persisted during these periods
by means of dormant seeds. It is hard to imagine a population in which densityindependent forces have no effect on population density or dynamics. However,
while density-independent mortality sources can limit the size of a population, they
cannot regulate it (Watkinson 1997).
43
10,000
1,000
100
10
Seeds
Seedlings
Population density
500
10
15
20
census number
Flowering
plants
25
30
Seeds
produced
Densitydependent
mortality
50
50
Densitydependent
fecundity
Population density
Fig. 14 Schematic diagram illustrating the role of density-dependent forces in regulating population density (Figure reprinted from Silvertown and Charlesworth 2001)
less vulnerable to infection in a population where individuals are widely spaced than
in one where trees are crowded and sunlight or nutrients are in short supply. Thus
mortality due to fungal attack may be density dependent. Finally, when the number of
adult plants is small, each individual will grow larger and produce more seeds than
when individuals are denser (Fig. 14). All these forces tend to dampen variations in
population density and thus to regulate population numbers.
Because so many plant populations appear to be regulated in some way, the
existence of density dependence has been investigated in a wide range of species.
Both observational and experimental approaches have been used. Two kinds of
observational studies have provided evidence for density-dependent population
regulation. First, ecologists have looked for positive correlations between plant
size and interplant distance, considering such patterns to be evidence that plant size
is controlled, to some degree, by the intensity of competition with neighbors. Other
kinds of observational studies have taken advantage of natural variation in population density, either in time or in space, to determine whether and how a populations
birth and death rates vary with density. However, tightly regulated populations are
1000
Fig. 15 Relationships
between original density of
seeds or plants and the final
yield biomass (in g) for
clover, Trifolium
subterraneum (top), and for a
grass, Bromus unioloides
(bottom) (Figure from
Harper 1977)
P. Bierzychudek
Total yield (g)
44
0 2.5
12.5
25
Seeds x 103/m2
N3
80
60
40
N2
20
0
N1
01 3 6
12
25
Density (plants/pot)
50
expected to exhibit little natural variation in density; thus the stronger the regulation, the harder it is to detect (Silvertown and Charlesworth 2001). Another
shortcoming of both kinds of observational studies is that spatial variation in
environmental factors could complicate the interpretation of observed trends
(Antonovics and Levin 1980). An alternative approach has been to alter density
experimentally, either in the field or in the greenhouse, and to measure how survival
and fecundity rates vary with density.
A large number of such studies have repeatedly demonstrated that variation in
population density can have dramatic effects on individual growth rates, fecundity
rates, and mortality rates (Harper 1977; Antonovics and Levin 1980). At relatively
low densities, individual plants tend to exhibit few reductions in performance.
However, at medium densities, reductions are often seen in growth rate and
reproductive output. Finally, at relatively high densities, mortality rates can
dramatically increase. For example, studies of how final biomass depends on the
density of seeds originally sown have repeatedly confirmed the law of constant
final yield (Fig. 15). Similarly, the relationship between plant weight and
plant density represented by the 3/2 self-thinning law (Eq. 1) illustrates the
powerful influence of density. Because these reductions are observed even in
controlled environments where herbivores and parasites are absent, it is clear that
these reductions are very often a consequence of resource competition among
conspecific neighbors.
The potential effect of intraspecific competition can be incorporated into the
previous model of population growth, shown here in the form of a difference
equation (contrast this with Eq. 3):
Nt
Nt1
6
1 aNt
In this so-called logistic model, a equals ( 1)/K, where K is the carrying
capacity of the environment for the species (in units of numbers of individuals).
45
3,500
number of individuals
3,000
2,500
2,000
1,500
07
20
08
20
09
20
10
06
20
05
20
04
20
03
20
02
20
01
20
20
20
00
1,000
year
This model differs from the geometric model only in its modification of the assumption that is a constant. The logistic model assumes that the growth rate is equal to
when Nt is near 0 and that it decreases linearly toward 1 as Nt approaches K. The
logistic model generates the population dynamics shown by the closed circles in
Fig. 16. A derivation of this model can be found in Begon et al. (1996).
Some readers may be more familiar with the continuous-time form of the logistic
model,
dN
KN
rN
7
dt
K
Equation 7 contains the same assumption about the linear dependence of the
population growth rate on N as does Eq. 6. Both models predict that a populations
numbers should grow until they reach an equilibrium size (K), at which point deaths
balance births. The observation in nature of a trajectory like that in Fig. 13 implies
that a populations dynamics are largely governed by intraspecific competition for
one or more limited resources. Many populations that initially display a pattern of
geometric growth eventually reach a more or less stable size like that predicted by
the logistic model.
Resource competition is not the only biotic interaction potentially capable of
regulating plant populations; interactions with enemies like herbivores, seed predators, and plant parasites such as fungi and bacteria also have the potential to act as
regulatory forces. An influential paper by Hairston et al. (1960) argued that the fact
that plants generally appear abundant and largely intact implied that it was
unlikely that plant populations could be regulated by their enemies. However, this
argument has been challenged for many reasons (see review by Crawley 1989).
Indeed, it is often assumed that natural enemies can regulate plant populations; for
example, efforts to use biological control to reduce weed populations are grounded
in this assumption (Halpern and Underwood 2006). In addition, a popular hypothesis
46
P. Bierzychudek
to explain why plant species that have been transported from their native location to
a new geographical region often become invasive is the enemy release hypothesis.
This hypothesis proposes that movement to a new location releases nonnative
species from the regulatory effects of the enemies that held them in check where
they were native.
The relative importance of natural enemies in regulating plant populations
remains controversial, however, because they have been less well investigated
experimentally. Much of the evidence supporting the role of natural enemies
comes from large-scale releases of herbivores for purposes of weed control; such
releases are neither randomized nor replicated. Better evidence comes from controlled experiments in which plants in plots protected from herbivore activity by
caging or insecticide application are compared to plants in unprotected control
plots. The results of such studies have been mixed, with vertebrate herbivores
typically exerting stronger regulatory effects than insects and some studies showing
no evidence for herbivore regulation (Crawley 1989). Because these methods of
herbivore exclusion have been shown to have unintentional treatment effects, even
those studies implicating herbivores as important do not necessarily provide compelling evidence for the role of natural enemies in regulating plant population
dynamics (Crawley 1989). Additionally, such studies are often limited to measuring
the impact of enemies on individual plant performance, and their results cannot
easily be scaled up to provide insights about the regulation of entire populations.
For example, a herbivore that reduces an individuals seed production might not
affect the populations dynamics if the availability of safe sites limits the numbers
of seeds that can germinate successfully (Crawley 1989; Halpern and Underwood
2006). Finally, studies investigating the effect of natural enemies on plant performance rarely investigate whether such effects are density dependent, as they must
be if they are to be able to regulate plant population dynamics (Halpern and
Underwood 2006). The role of natural enemies in regulating plant populations is
an important area in need of additional investigation, especially because the
findings of these efforts have important implications for the control of pests and
the management of plant invasions (Halpern and Underwood 2006).
47
vary from 1 year to the next, producing temporal patterns like that in Fig. 9. All
natural populations, regardless of their size, are influenced to some degree by
environmental stochasticity.
In contrast to environmental stochasticity, demographic stochasticity refers to
variation in vital rates arising from chance differences in the fates of different
individuals; this kind of variation arises from within the population itself rather than
from external forces. For example, an average plant in a population might be
expected to produce 100 seeds, but not every plant conforms to this average.
Some might make more than 100 seeds, some fewer. Demographic stochasticity
is primarily a concern for small populations, because in large populations, there
are abundant opportunities for these random deviations from the mean to cancel one
another out. For this reason, large populations are much more likely to follow the
law of averages. In a small population, however, it is likely that these random
interindividual differences will lead to deviations in the numbers of deaths or births
in different years and thus to a population size that varies randomly from 1 year to
the next. Since small populations also experience environmental stochasticity, they
can fluctuate in size to a considerable degree between years. This fluctuation is
important because it greatly increases their vulnerability to extinction.
The way environmental stochasticity affects population dynamics, and thus a
populations extinction risk, is important but somewhat counterintuitive. Temporal
fluctuations in vital rates do more than cause a populations dynamics to be more
variable over time; they can actually cause a population to grow more slowly than it
would in the absence of variability. Morris and Doak (2002) illustrate this effect
using the following example. Imagine a population of 100 individuals with an
annual growth rate, , that can take one of two values, 0.86 and 1.16, each value
occurring with a 50 % probability. The average of these two values is 1.01; thus, we
might reasonably expect that this population would have 14,477 individuals
500 years in the future:
100 1:01500 14, 477
However, the population will not grow at a rate of 1.01 every one of these
500 years. Each year, it will grow either at a rate of 1.16 or 0.86. If 1.16 in
exactly 250 years, and 0.86 in the other 250, which is quite probable, the population
would in fact have only 54 individuals 500 years from now, a huge difference from
the calculation in Eq. 8:
100 1:16250 0:86250 54:8
Of course other outcomes are possible in this probabilistic scenario, but this one
is the most likely. It is no accident that the computation that accounted for variation
in predicted a smaller population than the computation using the mean; incorporating stochasticity into models of population growth makes it likely that
populations will do worse than they would in a deterministic model (Morris and
Doak 2002).
48
P. Bierzychudek
In the preceding example, the simple average of the two values of , 1.01,
generated a very poor (and wildly overoptimistic) prediction of the populations
future dynamics. This simple average (the sum of n values divided by n) is also
known as the arithmetic mean. A less-familiar mean is the geometric mean (the nth
root of the product of n values). The geometric mean of 1.16 and 0.86 is 0.9988, and
using it instead of the arithmetic mean generates a more accurate prediction of the
populations growth rate in the face of environmental stochasticity:
100 0:9988500 54:8
10
The geometric mean of a series of numbers is always less than or equal to the
arithmetic mean. That the geometric mean yields a more accurate population
prediction should make sense, given that population growth is a multiplicative
process.
As this example illustrates, a population experiencing temporal variability in
vital rates might decline over time, even if in some years its growth rate, , is well
above 1.0. This fact has important implications for the persistence of species of
conservation concern. Using information about the amount of temporal variability a
population experiences, a prediction can be made about the likelihood that a
population will persist or go extinct within some specified time frame. Such
information can also be used to identify effective management options. These
investigations use a variety of modeling approaches collectively known as population viability analysis (PVA).
Over the last several decades, the development of models to assess the extinction
risk of threatened or endangered populations has been one of the most active areas
of research in plant (and animal) population dynamics. Morris et al. (1999) is an
excellent introduction to some of the most commonly used PVA approaches, and
Morris and Doak (2002) provide further elaboration; Brigham and Thomson (2003)
provide a good, brief overview. PVA models allow to vary over a range of values
from year to year, with that range representing the degree of environmental
variation a population experiences. Such models cannot forecast the future size of
the population with certainty; instead, they aim to forecast the probability that a
population will achieve a particular size (or become effectively extinct) by some
specified future time. The greater the interannual variability in population growth
rates, the greater the uncertainty associated with these forecasts.
To illustrate this approach, some of the data in Fig. 2 for the Hawaiian silversword
are analyzed here using the simple PVA for count data (i.e., unstructured data)
presented in Morris et al. (1999). The data come from 11 permanent plots that were
established on Mt. Haleakala in 1982 to permit long-term monitoring of the silversword population. All individuals in the plots were censused in 23 of the years
between 1982 and 2010 (Krushelnycky et al. 2013). The population in Fig. 2 shown
by the closed squares has fluctuated in numbers over the census period and since 2000
has appeared to be declining. What are the survival prospects for this population if
current trends continue? The first step in performing a count-based PVA is to estimate
values of , which is a stochastic version of the log of the population growth rate (see
49
1.00
cumulative probability of extinction
.80
.60
.40
.20
.00
0
100
200
300
400
500
time in years
Morris and Doak 2002 for details), and of 2, a measure of the stochastic variance in .
Morris et al. (1999) and Morris and Doak (2002) provide formulas for computing
these parameters. Following their procedure yields a value for of .001. The fact
that is negative means that the population will certainly go extinct; this is a
reasonable expectation given the population trend evident in Fig. 2. But how much
time will elapse before extinction occurs? To determine the likely time frame for this
event, the cumulative distribution function (CDF) of extinction probabilities can be
estimated (code for this computation is available in the R package popbio). To
estimate a CDF, it is important to define an extinction threshold, i.e., a number of
individuals below which the population becomes effectively extinct. In this example,
that threshold has been set to four individuals. The resulting CDF, shown in Fig. 17,
illustrates that without active management of some kind, this population of Hawaiian
silverswords is likely to be extinct within 200 years.
50
P. Bierzychudek
Table 1 A life table for the grass Poa annua, data from Law (1975), table adapted from Begon
et al. (1996)
x, age (in 3-month
periods in this
example)
0
1
2
3
4
5
6
7
8
ax, number of
individuals that live
to age x
843
722
527
316
144
54
15
3
0
lx, proportion
surviving to age
x
1.0
0.856
0.625
0.375
0.171
0.064
0.018
0.004
0
Structured models are based on the notion of a life table, a convenient way
to summarize demographic information for age-structured populations. First
developed for human populations, life tables contain information on how probabilities of survival and reproduction vary with an individuals age. A life table
summarizes data collected during repeated regular censuses of a cohort, which is
a group of individuals all born at the same point in time. This information can then
be used to calculate the cohorts (and, by extension, the populations) rate of
increase.
Each age is represented as a separate row in a life table (see Table 1), and
information on the survival and fecundity for each age is organized as a series of
columns. The first column of a life table contains the ages (x) of individuals in the
cohort, with x 0 representing the age of a newborn individual. (Because seeds
are so hard to observe, birth in plant life tables is often defined as the appearance
of a seedling.) While censuses are often conducted annually for organisms in
seasonal environments, census intervals may be chosen to be shorter (as in
Table 1) or longer than a year, depending on the life history of the organism.
The life table here is for an annual grass, Poa annua, and censuses were carried out
every three months.
At each census, the numbers of survivors of the cohort are counted. These data
are presented in the second column (ax). The original number of individuals in the
cohort, 843 in this example, is a0. These values can be used to compute each age
classs age-specific survivorship, lx (ax/a0), which is the proportion of the original
cohort that lives at least until age x. Age-specific fecundity, mx, is typically
quantified as the mean number of seeds (or seedlings) produced per individual
while it is age x.
The symbols used to represent these different vital rates are unfortunately not
standardized; some authors use Nx in place of ax or Bx in place of mx. Likewise,
survivorship (lx) is sometimes represented as the proportion of a cohort still alive, as
is the case here, and other times as a standardized number of survivors from
a hypothetical original cohort of 1,000. It is also worth noting that for organisms
51
with separate sexes, life tables are based on the number of female offspring
produced by a typical female, since the population growth rate in such species is
typically determined by the rate at which females reproduce. Since most plant
species are hermaphroditic, life tables for most plants need not make this
distinction.
The data in a life table can be used to compute the cohorts net reproductive rate,
R0, the average number of offspring that a typical individual produces over its
lifetime, i.e., per generation. The formula for R0 is
R0
k
X
lx mx
11
x0
where k is the final age used in the life table. Note that R0 differs from a simple sum
of the numbers of offspring produced at each age; it weights each reproductive
episode by the likelihood that an individual will live to that age. The units of R0 are
the expected numbers of offspring produced per newborn individual per generation.
In order to convert R0 to or to r, the generation time, G, must be computed, as
follows:
Xk
lx mx x
G Xx0
12
k
lm
x0 x x
The relationship between R0 and is then
1
R0 G
13
lnR0
G
14
It is important to note that life tables, like the simplest unstructured models
presented in section Causes of Different Temporal Patterns of Plant Population
Dynamics, assume that an individuals fecundity depends only on its age and is not
affected by population density. Thus a life table is implicitly a geometric growth
model. In that sense, it can accurately compute a populations current reproductive
rate, but it might do a poor job of forecasting future reproduction. Secondly,
because in many plants the correlation between age and size is not very strong
(Gurevitch et al. 2002), life tables are not appropriate tools for the study of many
plant populations; they are probably most appropriately applied to annual species,
as in Table 1. However, they provide a useful introduction to other kinds of
structured models.
Structured models of most plant species tend to use size classes rather than age
classes. The use of size classes introduces some complications into the modeling
process. In a life table, in which individuals are classified by their age, an
individual can have only two possible fates between successive censuses: it may
52
P. Bierzychudek
Seed
Seedling
1-leaf
2-leaf
3-leaf
4-leaf
Fig. 18 Life-cycle diagram for Panax quinquefolius, American ginseng. In this scheme, individuals are divided into six possible size classes. Information from annual censuses allows researchers
to estimate the probability of each of the transitions represented by the arrows (Reprinted from
Farrington et al. 2008)
move into the next age class, or it may die. When individuals are classified by size
rather than age, there are more possibilities. Between censuses, an individual may
(a) move from a smaller size class to one or more larger size classes (growth),
(b) move from a larger class to one or more smaller classes (regression),
(c) remain in the same class (stasis), or (d) die. These complex possibilities are
often displayed in the form of a life-cycle diagram. Figure 18 shows a life-cycle
diagram for American ginseng, Panax quinquefolius, an herbaceous perennial.
The arrows represent the possible changes that individual plants can undergo
between successive censuses, as well as the fact that plants having at least two
leaves can also produce seeds. Individuals that die between censuses are not shown
in the diagram.
To accommodate these complications, plant ecologists generally model a
structured populations dynamics with size-structured transition matrix models,
also known as Leslie matrix models, Lefkovitch models, or simply matrix
models. A transition is a period of time between successive population censuses,
during which individuals in the population may undergo changes in their status, like
those in Fig. 18. These models represent the populations status changes during
each of these transitions as a matrix of vital rates (Fig. 19). Each vital rage is
estimated from annual censuses of individually marked plants. A transition matrix
is square (i.e., it has equal numbers of rows and columns). There are as many
rows and columns as there are size classes. Each entry in the matrix has
two subscripts: the first (i) representing its row (i.e., the class it has transitioned
to) and the second (j) representing its column (i.e., the class it has transitioned
from). Each entry in the matrix, aij, represents the proportion of individuals
originally present in class j that transitioned to class i between the first and second
census.
M
0
0 .125
.601 .091
0
.011 .633 .82
53
From class (at time t):
1
a11
2
a12
3
a13
4
a14
a21
a22
a23
a24
a31
a32
a33
a34
a41
a42
a43
a44
1
To class
(at time t+1):
nt
15
30
100
nt+1
(0 15)
(0 30)
(.125 100)
(0 100)
12.5
11.745
101.155
Fig. 20 This example represents a population divided into three size classes. At time t there were
15 class-1 individuals, 30 class-2 individuals, and 100 class-3 individuals. Multiplication of the
matrix M by the vector nt as shown produces a new vector, nt+1, of 12.5 class-1 individuals, 11.745
class-2 individuals, and 101.155 class-3 individuals (since a fractional individual cannot exist,
these are often rounded to the nearest whole number)
ik
X
aij
15
i1
Conventionally, the first class in a transition matrix represents newborn individuals (i.e., individuals present at the second census that were not present at the first),
so the entries in the top row of the matrix are zero until reproduction has been
incorporated. The reproductive contribution of class j is defined as the mean
number of class-1 individuals present at time t + 1 that were produced between
the first and second censuses by individuals in class j at time t. Morris et al. (1999)
and Morris and Doak (2002) provide clear accounts of how to construct a transition
matrix from census data.
Figure 20 shows an example of a matrix (M) for a hypothetical plant population
in which individuals can belong to any of three size classes. In this example, these
transitions are possible: class-1 individuals can grow to class 2 or to class 3 or die;
class-2 individuals can stay in class 2, grow to class 3, or die; and class-3 individuals can stay in class 3 or die. Only class-3 individuals can reproduce. Figure 20
also shows two vectors (columns of numbers). These vectors represent the
populations size structure, i.e., the numbers of individuals present in each size
class at some particular census period. The sum of these numbers equals nt, the total
number of individuals in the population at time t.
Matrix models place vital rate data into a matrix format so that the operations of
matrix algebra can be used to project the populations size structure into the future,
54
P. Bierzychudek
55
56
P. Bierzychudek
et al. (1998) with data on vital rates in the burned population during the year of the
burn as well as 1 and 2 years afterward.
Gross et al. (1998) performed both a deterministic analysis as well as one that
incorporated stochastic variability by treating each vital rate as a random variable.
In the deterministic analysis, they created different matrices that represented
populations subject to one of three levels of trampling (no reduction from current
levels and 50 % and 100 % reductions of trampling mortality) in non-burn, burn,
and postburn years. By multiplying different matrices together, they created product matrices that simulated a variety of burn scenarios (e.g., burning every other
year, every 5th year, every 10th year) in combination with any of the three
trampling scenarios and computed for each one. In their stochastic analysis,
they explored 39 different management strategies, consisting of the three different
trampling levels combined with 13 different burn scenarios, ranging from no
burning to control burns carried out at intervals of between 1 and 20 years.
The studys results demonstrated that neither management strategy by itself was
sufficient to reverse the decline of H. montana (Fig. 21). However, they found that
population growth ( > 1) was possible if burning was combined with the elimination of some or all of the trampling. While one burn every 68 years was
predicted to maximize H. montanas growth rate, Gross et al. (1998) found that
decreasing the burn frequency to as much as once every 1216 years would still
allow the numbers of this threatened plant to increase. The stochastic analysis
produced a somewhat more optimistic outlook (compare Fig. 21) than the deterministic one. This finding runs counter to the idea described in section The Role of
Stochastic Influences, Especially in Small Populations that incorporating environmental variability often leads to forecasts of slower population growth. This result
could be due to the nature of the variability in this particular example or to negative
correlations in the variability of different vital rates (Doak et al. 2005).
Gross et al. (1998) asked what strategies would be effective in reversing
H. montanas observed decline. The same data can be used to carry out a PVA.
The goal of such an analysis is to forecast the probability of extinction if no
management were implemented. Morris and Doak (2002) reanalyzed Gross
et al.s (1998) data to produce such a forecast. Incorporating environmental variability by using a matrix-selection approach, Morris and Doak (2002) computed the
cumulative probability of extinction (which they defined as the populations falling
below 500 individuals, since most of the individuals are dormant seeds in the soil)
as a function of time. They found that, in the absence of any management action, the
population has nearly a 50 % probability of extinction within 50 years (Fig. 22).
Methods for these and other analyses using matrix models can be found in Caswell
(2001) and in Morris and Doak (2002), and code for carrying them out is available
in the R package popbio.
The incorporation of environmental variability is not the only important concern
when using matrix models. Another assumption of matrix models is that the
population has attained a stable size distribution. Until this occurs, the actual
population growth rate can be quite different and either larger than or smaller
than . A population in a highly variable environment may not have the opportunity
Elimination of trampling
50% trampling reduction
No trampling reduction
1.01
1
= 0.9845
0.98
= 0.9749
0.97
0.96
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
b 1.02
Mean population growth rate ()
Declining
population
= 0.9943
0.99
1.01
= 1.0038
Growing
population
a 1.02
57
Growing
population
1
= 0.9941
0.99
0.98
= 0.9845
Declining
population
0.97
0.96
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Length of burn cycle (years)
Fig. 21 The annual population growth rate, , of Hudsonia montana as a function of simulated
burn cycle length and level of trampling reduction for deterministic (a) and stochastic (b)
transition matrix models. Dashed lines in a represent the population growth expected while the
population achieves a stable size distribution
58
.50
cumulative probability of extinction
P. Bierzychudek
.40
.30
.20
.10
.00
5
10
15
20
25
30
35
40
45
50
changing in size very rapidly; therefore, the geometric model may often be an
appropriate one. For species of conservation concern, whose population sizes are by
definition well below K, the assumption of a lack of density dependence is certainly
appropriate, justifying the widespread use of these models for this purpose. Nevertheless, it is clear that there are some kinds of plant populations for which this
density-independent approach is unsuitable. For this reason, density-dependent
versions of matrix models have been developed (Caswell 2001: Morris and
Doak 2002).
The widespread use of matrix models, coupled with an appreciation of their
limitations/assumptions, has raised questions about their value and applicability.
Crone et al. (2013) used long-term data from 20 plant species to compare
the forecasts of matrix models for these species with their observed population
dynamics. They concluded that matrix models provided a good integration of a
populations vital rates during the time period during which those vital rates had
been estimated and that was indeed a suitable way to assess a populations status
and to evaluate management options. However, they found that in many instances,
matrix models failed to accurately forecast future population sizes. In evaluating
the possible causes of this failure, Crone et al. (2013) ruled out density dependence
and shortcomings in the number of sampled plants or census years, two often-cited
concerns about matrix models.
Instead, they concluded that the most plausible explanation for why matrix
models sometimes fail to accurately forecast future population behavior is that
the assumption of environmental constancy (even allowing for stochastic variation
about some mean) is not met (Crone et al. 2013). Especially in the face of the
environmental changes in temperature and precipitation currently occurring as a
result of anthropogenically increased levels of atmospheric CO2, it is clearly
desirable to develop ways to incorporate the likely effects of directionally changing
environmental parameters into models of population dynamics.
59
60
P. Bierzychudek
by human activities like urban and agricultural development, forest harvesting, etc.
Such anthropogenic habitat fragmentation has been recognized as one of the greatest
threats to species diversity. Regardless of the cause of patchiness, many plant species
are distributed within discrete patches of suitable habitat embedded in an unsuitable
habitat matrix; these patches can be connected by the dispersal of seeds and/or pollen.
Understanding the persistence of species in fragmented habitats often requires
adopting a spatial perspective that includes more than a single local population.
Regional assemblages of populations of the same species can take many forms.
The best-studied type of regional population assemblage is the metapopulation.
A metapopulation is a network of relatively small, local subpopulations connected
by migration. Because of their small size, individual subpopulations within the
larger metapopulation are prone to local extinction. Metapopulation theory has led
to the conclusion that in order for a metapopulation to persist over the long term,
there must be asynchronous, reciprocal dispersal between existing subpopulations
and from existing subpopulations to unoccupied patches of suitable habitat and that
the density of suitable habitat patches must exceed some threshold (Freckleton and
Watkinson 2002). The dynamics of the entire metapopulation are determined by
these processes of extinction, dispersal, and recolonization and thus are not a simple
function of the collective dynamics of local populations (Freckleton and Watkinson
2002). Likewise, the dynamics of local populations that are part of a metapopulation
cannot be completely understood without adopting a metapopulation perspective.
While metapopulation theory has had a strong influence on how animal
populations are studied, there are limited numbers of studies of plant populations
that take a metapopulation perspective, in part because the existence of seed
dormancy in many plant species complicates the quantification of extinction rates
(Husband and Barrett, 1996) and also because it is difficult to recognize what
constitutes a suitable habitat patch when it is unoccupied (Freckleton and
Watkinson 2002). Another way in which regional assemblages of plant populations
may differ from those of animals is that plants and their propagules are often very
long-lived, and their dispersal abilities are more limited than those of animals; thus
processes such as extinction and colonization may take place on much longer time
scales. Consequently, few studies have attempted to measure colonization, extinction, and recolonization rates and the density of suitable habitat patches for regional
assemblages of plant species (Freckleton and Watkinson 2002; Ouborg and
Eriksson 2004). In fact, the very applicability of the metapopulation concept to
plant species continues to be the topic of vigorous debate (Husband and Barrett,
1996; Freckleton and Watkinson 2002; Ouborg and Eriksson 2004).
Determining whether a particular plant species has a true metapopulation structure is more than an academic concern; it has important implications for how
species conservation should be approached. For species that exist as
metapopulations, it is inevitable that local populations will go extinct, so conservation efforts must not only protect existing subpopulations; they must also protect
unoccupied but suitable habitat and conserve dispersal opportunities (e.g., through
the creation of corridors). This is not necessary when local processes dominate
spatial dynamics. In addition to metapopulations, ecologists recognize other kinds
61
Censusing Populations
In the beginning of this chapter, repeated censuses were described as being at the
heart of studies of plant population dynamics. Of course, annual censuses must be
made at approximately the same time each year. Some studies of population
62
P. Bierzychudek
dynamics (known as count-based studies) only require information about how the
numbers of individuals in the population change over time. For these studies, it is
not necessary to know how each individual plants status changes temporally and
thus marking plants individually is unnecessary. It is not even necessary to count
the numbers of seeds in the soil, because such censuses are useful as long as they
represent counts of a constant fraction of the population each year (Morris
and Doak 2002). If a population is at or near a stable size distribution (see
section Incorporating Population Structure into Models and Analyses), this
assumption is likely to be met and seeds can be ignored. However, careful records
do need to be kept about the location of population boundaries, so repeated counts
can be made in the same area. Count data are the easiest data to acquire and are the
kinds of data most often collected by land managers responsible for monitoring
sensitive species. Analysis of these data is done by means of unstructured models
(see section The Role of Stochastic Influences, Especially in Small Populations).
However, it is relatively easy to track changes in the status of individual plants
over time and thus to go beyond count-based studies to incorporate information on a
populations age or size structure and how it changes over time. [A video by plant
ecologist James McGraw demonstrates some of these techniques using wild ginseng, Panax quinquefolium. http://www.youtube.com/watch?vu3CxPUr6cy4.]
These data can then be used to parameterize structured models (see section
Incorporating Population Structure into Models and Analyses). Gathering such
information typically requires marking each individual in the population (or a
randomly chosen subset of individuals) with a unique number, usually by attaching
numbered metal tags to the plants or inserting them into the ground nearby. A metal
detector can be a useful tool for relocating buried tags. Alternatively, for very
small plants, the corners of small sampling plots can be marked with nails and a
pantograph, photograph, or other method used to locate and relocate particular
individuals within the plot. However, rhizomatous plants and those whose position
may be altered by burrowing animals or by frost heaving can move a surprising
amount from 1 year to the next, making reliable re-identification difficult.
For structured population studies, decisions must be made about how to demarcate size classes or stages. This decision is partly based on convenience and
feasibility, but it is also important to find a reasonable compromise between
creating too few and too many classes. The more individuals in each class, the
more accurately their vital rates can be estimated. But the wider the boundaries of
the class, the more likely it is that the class will pool individuals of widely varying
sizes, with divergent demographic fates. See Caswell (2001) and Morris and Doak
(2002) for detailed advice about defining size class boundaries.
While most size classes are relatively easy to recognize, others are more
problematic. Some perennial plants have underground corms or other perennating
organs that, though alive, may remain dormant for one or more growing seasons.
Distinguishing dormancy from mortality requires multiple census years. Accurately
estimating individual fecundity can be difficult without repeated visits to a
population at the time of seed production, and many species have seeds that
remain dormant in the seed bank for anywhere from a few months to many years.
63
Future Directions
Transition matrix models will continue to be an important way to study the
dynamics of plant populations and to guide management decisions. Every year
these models grow increasingly sophisticated (Salguero-Gomez and de Kroon
2010). Some of the newest developments include ways to represent networks of
populations connected by dispersal, investigate the importance of ecological drivers
of population dynamics, explore the transient dynamics of populations responding
to changing conditions, and make better population forecasts in the face of temporal
and spatial stochasticity.
Understanding the effects of climate change on plant population dynamics, in
particular, is an area of high priority. Climate change is a long-term, uncontrolled
experiment whose effects on population dynamics are of great scientific and
practical importance. The large numbers of published studies making use of matrix
models facilitate the asking of questions such as: can we make robust predictions
about whether species in particular habitats or with particular life histories are more
or less vulnerable to the effects of stochasticity or climate change than others?
In the study of population dynamics in general, advances in molecular technologies are making it possible to identify and quantify soil microorganisms, permitting researchers to begin to explore how interactions with soil biota determine plant
population dynamics (Bever et al. 1997). And there are growing links between the
study of population dynamics and other biological subdisciplines, such as community ecology, ecosystem ecology, and ecophysiology, with the goal of providing a
greater mechanistic understanding of the processes underlying population dynamics and a better understanding of large-scale ecological processes.
References
Antonovics J, Levin DA. The ecological and genetic consequences of density-dependent regulation in plants. Annu Rev Ecol Systemat. 1980;11:41152.
Begon M, Mortimer M, Thompson DJ. Population ecology, a unified study of animals and plants.
3rd ed. Oxford: Blackwell; 1996.
Beschta RL. Cottonwoods, elk, and wolves in the Lamar Valley of Yellowstone National Park.
Ecol Appl. 2003;13(5):1295309.
Bever JD, Westover KM, Antonovics J. Incorporating the soil community into plant population
dynamics: the utility of the feedback approach. J Ecol. 1997;85:56173.
Brigham CA, Thomson DM. Approaches to modeling population viability in plants: an overview.
In: Brigham CA, Schwartz MW, editors. Population viability in plants. Berlin: Springer; 2003.
p. 14571.
64
P. Bierzychudek
Caswell H. Matrix population models: construction, analysis and interpretation. 2nd ed.
Sunderland: Sinauer; 2001.
Chapin FS, McGraw JB, Shaver GR. Competition causes regular spacing of alder in Alaskan shrub
tundra. Oecologia. 1989;79:4126.
Crawley MJ. Insect herbivores and plant population dynamics. Annu Rev Entomol.
1989;34:53164.
Crone EE, Menges ES, Ellis MM, Bell T, Bierzychudek P, Ehrlen J, Kaye TN, Knight TM,
Lesica P, Morris WF, Oostermeijer G, Quintana-Ascencio PF, Stanley A, Ticktin T,
Valverde T, Williams JL. How do plant ecologists use matrix population models? Ecol Lett.
2011;14:18.
Crone EE, Ellis MM, Morris WF, Stanley A, Bell T, Bierzychudek P, Ehrlen J, Kaye TN, Knight
TM, Lesica P, Oostermeijer G, Quintana-Ascencio PF, Ticktin T, Valverde T, Williams JL,
Doak DF, Ganesan R, McEachern K, Thorpe AS, Menges ES. Ability of matrix models to
explain the past and predict the future of plant populations. Conserv Biol. 2013;27(5):96878.
Doak DF, Morris WF, Pfister C, Kendall BE, Bruna EM. Correctly estimating how environmental
stochasticity influences fitness and population growth. Am Nat. 2005;166(1):E1421.
Enright N, Ogden J. Applications of transition matrix models in forest dynamics: Araucaria in
Papua New Guinea and Nothofagus in New Zealand. Aust J Ecol. 1979;4:323.
Farrington SJ, Muzika R-M, Drees D, Knight TM. Interactive effects of harvest and deer herbivory
on the population dynamics of American ginseng. Conserv Biol. 2008;23(3):71928.
Freckleton RP, Watkinson AR. Large-scale spatial dynamics of plants: metapopulations, regional
ensembles and patchy populations. J Ecol. 2002;90(3):41934.
Gross K, Lockwood Jr JR, Frost CC, Morris WF. Modeling controlled burning and trampling
reduction for conservation of Hudsonia montana. Conserv Biol. 1998;12(6):1291301.
Gurevitch J, Scheiner SM, Fox GA. The ecology of plants. Sunderland: Sinauer; 2002.
Hairston NG, Smith FE, Slobodkin LB. Community structure, population control and competition.
Am Nat. 1960;94:4215.
Halpern SL, Underwood N. Approaches for testing herbivore effects on plant population dynamics. J Appl Ecol. 2006;43:9229.
Harper JL. A Darwinian approach to plant ecology. J Ecol. 1967;55:24770.
Harper JL. Population biology of plants. London: Academic; 1977.
Husband BC, Barrett S. A metapopulation perspective in plant population biology. J Ecol.
1996;84:4619.
Hutchings MJ. The structure of plant populations. In: Crawley MJ, editor. Plant ecology. 2nd
ed. Oxford: Blackwell; 1997. p. 32558.
Krushelnycky PD, Loope LL, Giambelluca TW, Starr F, Starr K, Drake DR, Taylor AD,
Robichaux RH. Climate-associated population declines reverse recovery and threaten future
of an iconic high-elevation plant. Glob Chang Biol. 2013;19:91122.
Morris WF, Doak DF. Quantitative conservation biology, theory and practice of population
viability analysis. Sunderland: Sinauer; 2002.
Morris WF, Doak D, Groom M, Kareiva P, Fieberg J, Gerber L, Murphy P, Thomson D. A
practical handbook for population viability analysis. The Nature Conservancy; Washington,
DC, 1999.
Ouborg NJ, Eriksson O. Toward a metapopulation concept for plants. In: Hanski I, Gaggiotti OE,
editors. Ecology, genetics, and evolution of metapopulations. Burlington: Elsevier Academic
Press; 2004. p. 44769.
Salguero-Gomez R, de Kroon H. Matrix projection models meet variation in the real world. J Ecol.
2010;98:2504.
Salguero-Gomez R, Siewert W, Casper BB, Tielborger K. A demographic approach to study
effects of climate change in desert plants. Philos Trans R Soc B Biol Sci. 2012;367:310014.
Schemske DW, Bierzychudek P. Evolution of flower color in the desert annual Linanthus parryae:
Wright revisited. Evolution. 2001;55(7):126982.
65
Silvertown J, Charlesworth D. Introduction to plant population biology. 4th ed. Oxford: Blackwell;
2001.
U.S. Fish and Wildlife Service. Recovery plan for the Maui plant cluster. Portland: U.S. Fish and
Wildlife Service; 1997.
Watkinson AR. Plant population dynamics. In: Crawley MJ, editor. Plant ecology. 2nd ed. Oxford:
Blackwell; 1997. p. 359400.
Williams JL, Ellis MM, Bricker MC, Brodie JF, Parsons EW. Distance to stable stage distribution
in plant populations and implications for near-term population projections. J Ecol.
2011;99:11718.
Winnie Jr JA. Predation risk, elk, and aspen: tests of a behaviorally mediated trophic cascade in the
greater Yellowstone ecosystem. Ecology. 2012;93(12):260014.
Further Reading
Brigham CA, Schwartz MW, editors. Population viability in plants: conservation, management,
and modeling of rare plants. Berlin: Springer; 2003.
Caswell H. Matrix population models. 2nd ed. Sunderland: Sinauer; 2001.
Gibson DJ. Methods in comparative plant population ecology. Oxford: Oxford University Press;
2002.
Gotelli NJ. A primer of ecology. Sunderland: Sinauer; 2008.
Contents
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A Brief History of the Development of Community Assembly Concepts . . . . . . . . . . . . . . . . . . . . .
Dispersal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Abiotic Filtering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Biotic Interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Relationship Between Community Assembly and Coexistence Theory . . . . . . . . . . . . . . . . . . . . . . .
Phylogenetic Patterns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Biogeography and the Build Up of Species Pools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Scale Dependence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Future Directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Pattern-to-Process Mapping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Coexistence Theory and Community Assembly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Methods for Multitrophic Interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
68
68
71
73
75
76
77
80
83
84
84
85
85
85
Abstract
67
68
In many plant groups, close relatives tend to share similar phenotypic traits.
Therefore, patterns of phylogenetic relatedness within a community can also
be used to make inferences about community assembly mechanisms.
As the community and the species pool can be defined at a number of
different spatial and temporal scales, community assembly patterns often
show strong scale dependence. In some cases, a single process can produce
contrasting phenotypic patterns at different scales of analysis, while in other
cases different processes may have stronger influences on community assembly at different scales.
Species pools are shaped by dispersal of lineages among biogeographic
regions, in situ speciation within regions, and extinction. The characteristics
of the species pool often persist in local community patterns.
Community assembly studies are often limited in the extent to which specific
mechanisms can be inferred from community pattern. Future work should
focus on improved models of competition and coexistence dynamics in
community assembly as well as methods for considering multitrophic
interactions.
Introduction
Community assembly is the study of the processes that shape the identity and
abundance of species within ecological communities. Central to most studies of
community assembly is the concept of a species pool that is larger in geographic
scope than the local community under study. The species pool contains potential
colonists of the community, and many studies in this area focus on developing an
understanding of the role of dispersal, responses to abiotic conditions, and biotic
interactions in shaping local assemblages. Thus, community assembly considers
both the ecological interactions that shape local communities and the evolutionary
and biogeographic processes that lead to variation in the diversity and composition
of species pools across the globe.
69
70
Fig. 1 Basic conceptual model of community assembly in terms of species functional traits
(phenotypes). Empirically, an ecologist can consider the local community in relation to the species
pool of potential colonists. Habitat filtering is often hypothesized to limit the range of traits that
can successfully survive and establish at a site, as well as sometimes shifting the mean value
relative to the species pool. Competition, in its earliest forms in community assembly theory, was
predicted to favor the coexistence of species that differed in resource use or requirements, reflected
here in their functional traits. In the example here, competitive exclusion leads to a local
community of species with trait values that are more dissimilar than species in the original species
pool. Additional community assembly processes are not shown. (After Woodward and Diament
(1991))
While some early community assembly studies focused on forbidden combinations of species, much of the later research quickly transitioned to a focus on
patterns of phenotypic traits of community members rather than on species identity
per se. In plant community ecology, the focus has often been on functional traits,
which are defined as aspects of the plant phenotype that are indicative of variation
in ecological strategies of resource use, growth, and distribution in relation to
environmental conditions (Westoby and Wright 2006). Plant functional traits that
are relevant to community assembly can be anatomical or morphological traits,
such as specific leaf area (m2 g 1 biomass), root depth, or seed size, or they can be
ecophysiological measures that reflect the integrated activities of several related
plant processes, such as maximum photosynthesis rate or photosynthetic water use
efficiency (mol CO2 assimilated mol 1 water transpired). As these functional traits
can be measured on most if not all plants within and across communities, they offer
a phenotypic common currency that can be used to draw generalizations across
species and to make inferences about the mechanisms that shape community
patterns. For example, in studies of succession, environmental sieves or filters are
often hypothesized to drive convergence or clustering in phenotypic traits (relative
to a null model), whereas competition patterns of the sort originally proposed by
Diamond are typically predicted to produce phenotypic overdispersion, where
co-occurring species are more dissimilar in traits than expected (Fig. 1; Weiher
and Keddy 1999). More recent community assembly studies have refined these
predictions in a number of ways, as discussed in subsequent sections.
A number of methodological considerations arise when sampling and analyzing
functional traits in a community assembly context. Functional traits may vary
71
Dispersal
Dispersal refers to the movement of an individual organism during its lifetime, from
its place of birth to the location where it produces offspring. As plants are sessile
organisms, in most species, dispersal only occurs once during the life cycle at the
seed stage. Once a plant germinates, it occupies a single location for the rest of its
life. In addition to this mechanism, a small number of species are able to disperse via
72
73
Abiotic Filtering
One of the central metaphors in community assembly is that of a habitat filter,
where the abiotic environment filters out species by limiting establishment or
survival at particular sites. As plant dispersal is often relatively undirected, seeds
may often arrive at locations where conditions are not favorable for germination or
long-term survival. These filters can impact plants at any life stage and can involve
any of a number of abiotic factors singly or in combination.
Many plant species have specific abiotic requirements for successful germination, and thus the germination stage represents the first point at which habitat
filtering can occur. Germination cues can include moisture, temperature, light,
photoperiod, and even fire or smoke in some species adapted to fire-prone environments. Many species require specific combinations of abiotic cues, such as a period
of cold temperature followed by a photoperiod indicative of long days. Reliance on
these cues can help to ensure that a species will not germinate and die in unfavorable conditions. Some species are able to persist in a dormant state as a seed for long
periods of time waiting for the proper cues to trigger germination, but the length of
time that seeds remain viable varies widely among species. The ability of some
species to persist for extended periods in the seed bank can complicate the task of
quantifying community membership at a particular site, and an examination of
74
the seed bank (and testing for seed viability) may be required to definitively
conclude that a species is absent from a site. This is most relevant in communities
that exhibit substantial variation in abiotic conditions over time, as different species
may use the same habitat at different times of the year or in different years,
depending on year-to-year variation in weather, remaining dormant in the seed
bank at other times.
Abiotic factors can also cause mortality or prevent successful reproduction at
any time during the life cycle from germination through reproductive maturity.
Species vary in requirements for light, nutrients, and water as well as in tolerance to
drought and temperature, and any of these factors can cause mortality at any stage.
An important consideration is that brief, extreme climatic events can have strong
impacts on species survival. For example, the average climatic conditions at a site
may be ideal for the growth and reproduction of a species, but a brief period of
extreme cold or heat or a short but severe drought that occurs infrequently can cause
significant mortality and effectively remove particular species from a site. For
example, a severe drought associated with an El Nino event in the 1980s is thought
to have had persistent and long-lasting impacts on the species composition, and
associated functional traits, of a tropical forest on Barro Colorado Island, Panama
(Feeley et al. 2011). Therefore, in considering the role of abiotic conditions in
filtering species from a site, it may be just as important to consider the variance or
the extremes of abiotic conditions as it is to consider the average values.
Practically speaking, it can be challenging to distinguish between habitat filtering and dispersal limitation when a species is completely absent from a site.
Simple experiments can be helpful in testing for habitat filtering. On the most
basic level, these experiments involve transplanting individuals either as adults or
as seeds to the site and monitoring germination and/or survival. In situations where
these experiments are impractical, seed traps or detailed examination of the seed
bank can be useful in ruling out dispersal limitation as the cause of a species
absence.
An important consequence of abiotic filtering is that species composition typically changes along environmental gradients. For example, there is widespread
evidence that plant communities change in predictable ways along gradients of
light, water availability, soil fertility, elevation and latitude, among other factors.
These changes in species identity are also often reflected in changes in the functional traits of species, such that average trait values across species in the community can shift along a gradient. For example, woody plant leaf functional traits
change consistently across a gradient of soil water availability in coastal California
and across microtopographic gradients in the Ecuadorian Amazon (Kraft
et al. 2008; Cornwell and Ackerly 2009). Another frequently documented pattern
is that the breadth or variance of strategies seen at any point along the gradient is
often smaller than is seen across the gradient as a whole. The significance of these
observations i.e., shifts in the mean of trait values and reduction in the range or
variance in trait values at points along a gradient is typically documented using a
null model approach, comparing observed communities to hypothetical communities assembled at random from the regional species pool.
75
Biotic Interactions
Just as abiotic factors can serve as filters to prevent establishment of species,
interactions between plants and other organisms can have important consequences
for community assembly. Competition and natural enemies (herbivores, parasites,
and pathogens) can reduce growth and survival of plants at a particular site, and
positive interactions can allow species to establish and persist at sites where they
would otherwise be unable to survive. In many conceptual models of community
assembly, biotic interactions are often considered to impact community assembly
after abiotic filtering has occurred. While this may be true if the primary habitat
filter occurs at the germination stage, in reality biotic and abiotic factors are
likely important throughout the lifecycle of most plants. Persistence in a community requires tolerance of stresses in the germination, establishment, and adult
reproductive phases, to ensure reproduction of the next generation.
As stated earlier, competition has long been considered to be a central biotic
factor in community assembly, dating back to Jared Diamonds initial study of bird
communities on islands (and before that back to Darwin, writing in the Origin of
Species). Competition is hypothesized to impact community assembly by the
failure of species to establish or persist at a location in the face of competitive
interactions. Early community assembly theory focused on the competitive exclusion principle (Hardin 1960), which hypothesizes that complete competitors
cannot coexist, meaning that species are more likely to be able to coexist if they
have niche differences. Early work in this area focused on the concept of limiting
similarity, which hypothesized that there was a finite limit to how similar two
coexisting species could be. While theoretical work has since suggested that there is
not likely to be an absolute limit to similarity, the general idea that differences
between species promote coexistence by reducing competition has persisted as a
central theme in many community assembly studies. To date, many plant community assembly studies have approached competition by documenting differences in
the niches or phenotypes of co-occurring species and testing whether those differences are greater than what might be expected by chance. For example,
co-occurring plants in sand dune plant communities in New Zealand and forests
in the Ecuadorian Amazon are often more phenotypically distinct from each other
than predicted by null models (Stubbs and Wilson 2004; Kraft et al. 2008). In many
ways, this approach has direct links to Jared Diamonds initial approach of
documenting forbidden combinations of species on islands. While phenotypic
patterns that are consistent with competition are regularly detected in plant communities, they are far less common than evidence for habitat filtering.
Herbivores, parasites, and pathogens, collectively referred to as natural enemies,
can also have important and wide-reaching consequences for community assembly.
One challenge in this area is that community assembly studies typically focus just on
members of one guild or functional type (e.g., trees or herbaceous plants) and often
have not considered other trophic levels. In some cases, the impact of natural enemies
can be studied primarily through plant distributions. For example, if species suffer
primarily from natural enemies that are species specific, seedlings growing near adult
76
trees of the same species should suffer more negative effects than seedlings growing
far from adults, as natural enemies can become concentrated near adult trees
(reviewed in Wright 2002). In this case, the study of plant distribution patterns within
communities can offer some insight into the role of natural enemies. However, in
other cases, we likely need improved conceptual models and approaches to effectively incorporate natural enemies into community assembly studies.
Positive interactions between species can also have profound impacts on community assembly, allowing species to establish or persist at sites where they would
otherwise be unable to survive. Many of these associations are between plants and
other organisms. For example, associations with mycorrhizae and nitrogen-fixing
bacteria allow many plant species to gain access to essential nutrients more
effectively, and many species rely on insect or animal pollinators for reproduction.
The absence of these mutualist partners can effectively exclude plants from particular sites. Our understanding of these relationships in a community assembly
context is hampered by the same limitations as our understanding of natural
enemies many studies typically focus just on plants, not on other groups within
a community. While it is possible to study some consequences of plant-pollinator
interactions primarily through the plant community (Sargent and Ackerly 2008),
new approaches will be needed to fully incorporate positive interactions that extend
beyond a single trophic or functional group into community assembly studies.
It is also well understood that plants can have positive effects on each other.
These impacts most commonly involve an amelioration of environmental stress
or a reduction in herbivore pressure via associational defenses. For example, in
hot and dry environments, some species are known to function as nurse plants
by modifying the nearby microclimate enough to allow other species to be able to
establish. However, many positive interactions between plants are known to be
highly context dependent. For example, in one globally replicated experimental
study, plants growing at lower elevations on mountains were often found to
compete with one another, while species growing at higher elevations on the
same mountains (which is presumably a more stressful environment) were found
to have positive effects on one another (Callaway et al. 2002). These findings
highlight that most positive interactions (and perhaps many species interactions
in general) typically include both a positive and a negative component and that
the relative importance of these components for community assembly can shift as
abiotic conditions change. This also highlights a general but understudied challenge within the topic of community assembly disentangling the interactions
among abiotic and biotic filters.
77
Phylogenetic Patterns
In a famous quote from the Origin of Species (1859), Darwin noted that
As the species of the same genus usually have. . .much similarity in habits
and constitution, . . . the struggle will generally be more severe between them, if
78
they come into competition with each other, than between the species of distinct
genera. His observation reflected the general knowledge of any experienced
systematist or field naturalist that related species tend to be ecologically similar;
e.g., one would expect two grass-eating rabbit species to compete directly for the
same food sources, whereas a grain-eating mouse and a carnivorous fox are
utilizing quite different resources. In the first half of the twentieth century, experimental studies of competition by Gause and the development of Lotka-Volterra
competition theory led to the development of the competitive exclusion principle
(Hardin 1960), discussed earlier, which posits that species competing for the same
resources could not coexist in a community. Putting these ideas together, ecologists
in the mid-twentieth century suggested that species of the same genus would not
live together in local communities, at least not as often as one might expect if
communities were assembled randomly from the available species in a regional
species pool. This prediction was supported in studies of animal communities on
islands, compared to the fauna of adjacent mainland regions. These studies provide
some of the earliest examples of null models in ecology, discussed above.
Starting in the 1960s, the study of phylogenetics was revolutionized by conceptual, computational, and empirical advances, most notably the breakthroughs in
molecular biology leading to the modern era of DNA sequencing. With highresolution, well-supported phylogenies available, new methods have been developed to reexamine classical questions in ecology and evolutionary biology.
The study of plant communities presented particular challenges, as the deeper
structure of the angiosperm phylogeny had never been well understood and molecular data brought a number of surprises. The first breakthroughs came in the 1990s,
quickly leading to a broad community effort under the Angiosperm Phylogeny
Group and a rapidly growing consensus about major patterns in flowering plant
phylogeny. Plant ecologists moved quickly to utilize the newly available phylogenies to tackle large-scale problems in adaptive evolution, diversity, and community assembly (Webb 2000). Molecular data also provide branch lengths that
quantify the degree of relatedness among species, and fossil calibrations can be
applied to estimate branch lengths in millions of years since species diverged from
their most recent common ancestor. The phylogenetic distance between two species
is defined as the distance from one species down the phylogeny to the common
ancestor and back up to the other species (in other words, two times the age of their
most recent common ancestor).
The phylogenetic structure of a community can be described in a number of
ways, using quantitative metrics based on the phylogenetic relationships for the
community, pruned from the larger phylogeny of all plants. As in the examples
discussed above, statistical analyses of phylogenetic community structure consider
a local community relative to a null model of communities assembled from a
regional species pool. Two simple measures of phylogenetic community structure
are the mean phylogenetic distance, defined as the average of the phylogenetic
distances between all pairs of taxa in a community, and the mean nearest neighbor
distance, defined as the average distance from each species to its closest relative.
Using these measures, the net relatedness index (NRI) and nearest taxon index (NTI)
79
80
81
Fig. 2 Phylogenetic community structure of oak-dominated communities in Florida, demonstrating phylogenetic overdispersion within each of the three habitat types. Oaks within each of the
three major phylogenetic lineages occur in each community, and null model analysis reveals that
this pattern is not expected by chance (Redrawn with permission from Cavender-Bares
et al. (2004b))
82
83
tolerances, so close relatives spread out across the landscape and occupy different
habitats. Improved knowledge of phylogenetic history, the fossil record, and
the climatic history of different regions of the world will continue to shed light
on these fundamental questions in the evolution of regional floras and their influence on the assembly of local communities.
Scale Dependence
Community assembly studies typically focus on comparing the members of a focal
community to a regional pool of potential colonists. While this step might seem
clear in theory, in practice the definition of an appropriate boundary for a community and a species pool is fraught with uncertainty. It is often best to simply
acknowledge that there will be several possible ways in which to delineate the
community and the species pool and that each combination may reflect the action of
different assembly processes. For example, a species pool could be defined as any
species in the vicinity of the focal community that might be able, based on known
dispersal distances, to disperse a seed into the community within 1 year or one
generation of the focal species. A broader-scale analysis might consider the species
pool to be any species in the region, even if it would likely take longer than one
generation for some species in the pool to disperse into the local community. It is
essential to include an understanding then of how the pool and community were
defined when drawing conclusions based on community assembly analyses an
analysis based on a narrowly defined species pool might only be appropriate for
making inferences about short-term ecological processes, whereas an analysis
based on more broadly defined pool could reflect the action of processes acting
over multiple generations.
With this scale dependence in mind, there are a number of cases where a single
ecological process is predicted to produce contrasting patterns depending on the
scale of analysis. For example, a narrowly defined community sample at a small
spatial scale that only contains a single habitat type might readily demonstrate
phenotypic clustering or other patterns consistent with habitat filtering when
compared to a broader species pool that contains multiple habitat types. But if the
community sample is broadened to include two or more habitat types in the same
sample, it is conceivable that new, larger-scale analysis will reveal overdispersion,
reflecting the aggregation of two or more distinct phenotypic clusters of species
that are different from each other (Fig. 3). In this case, a single ecological
phenomenon environmental filtering will produce different phenotypic dispersion patterns depending on the scale of analysis.
In summary, it is essential for researchers to be cognizant of the criteria that are
used to delineate a community and a species pool in a community assembly analysis
and also to recognize that any inferences from the analysis will be conditioned on
those criteria, as patterns will likely shift as the scope of the pool and community is
altered. When possible, explicitly varying the scope of the pool and the sample can
be used to detect the action of processes operating at different spatial scales.
84
Fig. 3 A single community assembly process (habitat filtering into specific microsites) can
produce contrasting phenotypic patterns depending on the scale of analysis. If a microsite is
compared to the species pool, either microsite will show low dispersion (phenotypic clustering).
However, if a large spatial scale is used to define the community that includes both microsites, this
new larger community sample across sites may reveal higher phenotypic dispersion than expected
when compared to the species pool
Future Directions
The study of community assembly theory has seen considerable development since it
was pioneered in the middle of the twentieth century. However, considerable challenges remain, and here we highlight three essential areas where additional work is
needed, including better methods to distinguish between multiple processes in producing community patterns, a more complete incorporation of coexistence theory into
community assembly studies, and better consideration of multitrophic interactions.
Pattern-to-Process Mapping
Early work in community assembly focused on phenotypic convergence driven by
abiotic filters and phenotypic disparity driven by competitive exclusion. However,
it has become increasingly apparent that many community assembly processes can
produce similar patterns. For example, high phenotypic similarity of species within
a community can be produced by habitat filtering, in situ speciation, or pollinator
facilitation (Emerson and Gillespie 2008; Sargent and Ackerly 2008), among other
processes (Cavender-Bares et al. 2009). In some cases, additional information such
as the timescale of the analyses (which is implicit in the spatial and temporal criteria
used to define the species pool and the community) or the pollination syndromes of
the species in the community can be used to distinguish between potential mechanisms, but in other cases, definitive links between process and pattern can be
85
References
Adler PB, Ellner SP, Levine JM. Coexistence of perennial plants: an embarrassment of niches.
Ecol Lett. 2010;13:101929.
Callaway RM, Brooker RW, Choler P, Kikvidze Z, Lortie CJ, Michalet R, Paolini L, et al. Positive
interactions among alpine plants increase with stress. Nature. 2002;417:8448.
Cavender-Bares J, Kozak KH, Fine PVA, Kembel SW. The merging of community ecology and
phylogenetic biology. Ecol Lett. 2009;12:693715.
86
87
Webb CO. Exploring the phylogenetic structure of ecological communities: an example for rain
forest trees. Am Nat. 2000;156:14555.
Weiher E, Keddy PA. Ecological assembly rules: perspectives, advances, retreats. Cambridge:
Cambridge University Press; 1999.
Westoby M, Wright IJ. Land-plant ecology on the basis of functional traits. Trends Ecol Evol.
2006;21:2618.
Woodward FI, Diament AD. Functional approaches to predicting the ecological effects of global
change. Funct Ecol. 1991;5:20212.
Wright SJ. Plant diversity in tropical forests: a review of mechanisms of species coexistence.
Oecologia. 2002;130:114.
Further Reading
Abrams P. The theory of limiting similarity. Annu Rev Ecol Syst. 1983;14:35976.
Ackerly DD. Adaptation, niche conservatism, and convergence: comparative studies of leaf
evolution in the California chaparral. Am Nat. 2004;163:65471.
Ackerly DD, Cornwell WK. A trait-based approach to community assembly: partitioning of
species trait values into within- and among-community components. Ecol Lett.
2007;10:13545.
Bertness MD, Callaway R. Positive interactions in communities. Trends Ecol Evol. 1994;9:1913.
Bruno JF, Stachowicz JJ, Bertness MD. Inclusion of facilitation into ecological theory. Trends
Ecol Evol. 2003;18:11925.
Cavender-Bares J, Ackerly DD, Baum DA, Bazzaz FA. Phylogenetic overdispersion in Floridian
oak communities. Am Nat. 2004;163:82343.
Cavender-Bares J, Kitajima K, Bazzaz FA. Multiple trait associations in relation.to habitat
differentiation among 17 Floridian oak species. Ecol Monogr. 2004;74:63562.
Cornelissen JHC, Lavorel S, Garnier E, Diaz S, Buchmann N, Gurvich DE, Reich PB, et al. A
handbook of protocols for standardised and easy measurement of plant functional traits
worldwide. Aust J Bot. 2003;51:33580.
Cornwell WK, Schwilk DW, Ackerly DD. A trait-based test for habitat filtering: convex hull
volume. Ecology. 2006;87:146571.
Diaz S, Noy-Meir I, Cabido N. Can grazing response of herbaceous plants be predicted from
simple vegetative traits? J Appl Ecol. 2001;38:497508.
Fine PVA, Mesones I, Coley PD. Herbivores promote habitat specialization by trees in Amazonian
forests. Science. 2004;305:6635.
Gilbert GS, Webb CO. Phylogenetic signal in plant pathogen-host range. Proc Natl Acad Sci USA.
2007;104:497983.
Kraft NJB, Ackerly DD. Functional trait and phylogenetic tests of community assembly across
spatial scales in an Amazonian forest. Ecol Monogr. 2010;80:40122.
McGill BJ, Enquist BJ, Weiher E, Westoby M. Rebuilding community ecology from functional
traits. Trends Ecol Evol. 2006;21:17885.
Paine CET, Baraloto C, Chave J, Herault B. Functional traits of individual trees reveal ecological
constraints on community assembly in tropical rain forests. Oikos. 2011;120:7207.
Ricklefs RE. Community diversity relative roles of local and regional processes. Science.
1987;235:16771.
Ricklefs RE, Schluter D. Species diversity in ecological communities: historical and geographical
perspectives. Chicago: University of Chicago Press; 1993.
Ricklefs RE, Travis J. A morphological approach to the study of avian community organization.
Auk. 1980;97:32138.
Shipley B. From plant traits to vegetation structure: chance and selection in the assembly of
ecological communities. Cambridge: Cambridge University Press; 2009.
88
Swenson NG, Enquist BJ. Opposing assembly mechanisms in a Neotropical dry forest: implications for phylogenetic and functional community ecology. Ecology. 2009;90:216170.
Swenson NG, Enquist BJ, Pither J, Thompson J, Zimmerman JK. The problem and promise of
scale dependency in community phylogenetics. Ecology. 2006;87:241824.
Valiente-Banuet A, Verdu A. Facilitation can increase the phylogenetic diversity of plant communities. Ecol Lett. 2007;10:102936.
Vamosi SM, Heard SB, Vamosi JC, Webb CO. Emerging patterns in the comparative analysis of
phylogenetic community structure. Mol Ecol. 2009;18:57292.
Webb CO, Ackerly DD, McPeek MA, Donoghue MJ. Phylogenies and community ecology. Annu
Rev Ecol Syst. 2002;33:475505.
Weiher E, Keddy PA. Assembly rules, null models, and trait dispersion new questions front old
patterns. Oikos. 1995;74:15964.
Contents
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Pollination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
How Do Plants Benefit? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
How Do Animals Benefit? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Dispersal Agents: Who Is Involved? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Precision of Delivery Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
How Plants Manipulate Fertilization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Genetic and Evolutionary Consequences of Pollination Patterns . . . . . . . . . . . . . . . . . . . . . . . . . .
Evolutionary Dynamics: Evolution in Action . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Pollination Is Just Part of the Story . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Dispersal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
How Do Plants Benefit? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
How Do Animals Benefit? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Seed Packaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Dispersal Agents: Who Is Involved? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Fruit Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Patterns of Dispersal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Evolutionary Dynamics: Evolution in action . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Seed Dispersal Is Just Part of the Story . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Synthesis and Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
The Systematics of Associations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Effects of Dispersal Patterns and Ecological Heterogeneity on
Genetic Organization of Populations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Pollination, Dispersal, and Human Activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Future Directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
90
91
91
91
92
92
96
98
100
101
102
102
102
103
103
105
106
109
109
110
110
110
112
114
115
115
Y. Linhart (*)
Department of Ecology and Evolutionary Biology, University of Colorado, Boulder, CO, USA
e-mail: yan.linhart@colorado.edu
# Springer Science+Business Media New York 2014
R.K. Monson (ed.), Ecology and the Environment, The Plant Sciences 8,
DOI 10.1007/978-1-4614-7501-9_21
89
90
Y. Linhart
Abstract
Introduction
Pollination events can make headline news. Seriously. Consider the headlines
below, and note the sources: Huffington Post, New York Magazine, Washington
Post. . .not bad for press coverage of a botanical topic.
Corpse Flower, Worlds Stinkiest Plant, Blooms In Washington At
U.S. Botanic Garden Huffington Post 7/21/2013
Washingtons Stinkiest Flower to Reach Peak Smell on Monday
New York Magazine 7/21/13
The corpse flower is in bloom Washington Post 7/22/13
Not to be outdone, fruits can also get press coverage. The ginkgo tree produces
fruit that smells so foul that stories in the New York Times (November 5, 2010),
Washington Post (October 10, 2009), and other newspapers (e.g., Chicago Tribune,
Orange County Register, Toronto Sun) have commented on the unhappiness of
residents of neighborhoods where the fruit-producing trees have been planted, and
Washington DC horticulturists even tried to sterilize female ginkgos to avoid the
olfactory assault of ripening fruit. . .to no effect.
Plants are immobile and therefore depend on the actions of curious and hungry
strangers who will disperse their pollen and seeds. In some cases, the evolutionary
drive to attract pollen or seed dispersers gets pretty extreme as in the case of the
corpse flower and ginkgo, about which more later. But the excitement conveyed by
the news coverage at least for the corpse flower illustrates nicely why the topics
of pollination and dispersal provide entertainment and fascination. These topics are
also sources of scientific curiosity provided by the esthetics of colorful and odoriferous flowers and fruit and the impressive gymnastics that animals often go
through to get at rewards of pollen, nectar, fruit, and other goodies.
91
Pollination
How Do Plants Benefit?
Sex makes the world go round even for sessile plants, so that is what pollen
dispersal is about: improving the likelihood that individuals produce seeds that
will pass on their genes. It is true that under some specific circumstances,
certain plants are able to set seed without the intervention of animals or wind.
These plants are said to beautogamous or self-pollinating. However, given the
great preponderance of cross-pollination in the plant kingdom, clearly there are
significant advantages to mating with other individuals.
92
Y. Linhart
orchid bees use to mark territories. Some plants also produce various oils (used as
food) or resins (as nest-building materials) within their flowers. Some plants exploit
the passionate libido of male bees, wasps, or flies by mimicking the appearance and
sometimes the odor of females. These female look-alikes get mounted by males
whose exertions then either pick up pollen or deposit it (Gaskett 2011; Ellis and
Johnson 2010). Finally, other plants like the corpse flower take advantage of the
propensity of certain insects to be attracted to smelly corpses: they emit odors sure
to keep us away but wonderfully evocative to these carrion-feeding beetles and
flies, which are fooled by the smells and pollinate the flowers.
The various rewards offered by plants to attract pollinators are important drivers
of ecosystem function. For example, virtually all of the 25,000 or so bee species in
the world depend on pollen and/or nectar for their food. At least 650 species of birds
are obligate nectar feeders (about half of them are hummingbirds) and many more
such as orioles, warblers, and finches will partake periodically. About 10 % of bats
are pollinators for over 500 species of plants belonging to about 70 families.
93
who had never heard of the senders. Some precision is needed to ensure delivery
to appropriate locations. This is where plants can take advantage of the diversity
of animals that visit flowers and the diversity of sensory systems in these animals.
Plants have diversified greatly in terms of their mechanisms of attraction.
Figure 1 provides a small sample of the diversity of floral shapes, colors, and
architectures used to attract pollen dispersers, including the corpse flower.
As the animals approach the flowers, they perceive one or more of the following
features: (1) fragrance, (2) color, (3) morphology, and finally (4) rewards provided.
Also, the timing (day or night, and flowering season) of the rewards can be
manipulated to take advantage of the times of activity of their pollinators. Finally,
depending on the plant and pollinator involved, it seems that all sensory abilities of
pollinators can be exploited: sight, smell, taste, touch, and even hearing.
A combination of these features can attract specific pollinators and ensure reasonably accurate pollen delivery. For example, birds such as hummingbirds in the
Americas and sunbirds, sugarbirds, and honeyeaters on other continents are attracted
to red or orange flowers with little or no fragrance, which secrete large amounts of
dilute nectar, and are often long and tube shaped. Conversely, bees tend to be attracted
to flowers with strong fragrances, which are often yellow or blue and produce smaller
amounts of more concentrated nectar and lots of pollen (Proctor et al. 2012).
Some features of flowers are only visible under ultraviolet light: these patterns
can manifest themselves as lines or spots on petals and often serve the important
function of guiding pollinators to nectar rewards. As such these patterns are referred
to as nectar guides and are very poorly studied (Primack 1982).
At night, moths tend to gravitate towards flowers with delicate sweet smells like
jasmines that are white in color, while bats tend to visit greenish to purplish flowers
with strong smells of fermentation. These sets of floral characteristics are usually
called pollination syndromes (Proctor et al. 2012).
These descriptions do not imply that pollinators are fixed in their preferences and
will not visit flowers that deviate from these characteristics. For example, there are
multiple reports of hummingbirds visiting thistles, white-flowering jasmines, lavenders with intense blue flowers, and pink apple blossoms. Conversely, bees can
visit red or white flowers, while hawk moths will go to yellow or pink flowers. Such
behaviors indicate that pollinators have to be expedient in their choices: when their
preferred menus are not available, they make do. These patterns of pollinator
flexibility have led some students of pollination to doubt the accuracy of pollination
syndromes. More recent work has addressed this issue in a comprehensive manner
and has concluded that these patterns of preferences are valid in many situations,
but it is important to remember that many species have flowers whose signals
are understood by many animals and offer rewards accessible to many species
(Fenster et al. 2004).
Given that the signals broadcast by flowers often generate predictable behaviors
by specific visitors, it follows that when floral signals change, new visitors can be
attracted. Such shifts in visitor identity can change the patterns of pollen dispersal
from this plant. This matter is discussed in detail later in the section dealing with
evolution in action.
94
Y. Linhart
Fig. 1 Sample of diversity of flower colors and architectures used by plants to attract pollinators.
Top row: paperwhite, Narcissus papyraceus; cardoon, Cynara cardunculus. Second row: corpse
flower, Amorphophallus titanicum; dahlia, Dahlia sp.; saffron, Crocus sativus. Third row: sage,
Salvia cinnabarina; flamingo flower, Anthurium andraeanum; pincushion tree, Leucospermum
sp. Bottom row: tomato, Lycopersicon esculentum; violet, Viola sp. (Corpse flower photo from
Wikia; all other photos by YBL)
95
Touch and hearing have also been documented as factors relevant to some
specific pollinators. Certain bumblebees are especially fond of flowers such as
snapdragons, whose petals have rough surfaces, which enable the insects to grip
the flowers firmly and extract the nectar rewards more easily (Whitney et al. 2009).
Bats depend on sonar and hearing as they navigate their world, and at least two
plants are now known to help bats locate flowers and nectar sources by focusing
their hearing. In the tropical vine Marcgravia, there is a dish-shaped leaf positioned
so as to provide characteristic echo signatures that serve as a beacon towards the
open flowers (Simon et al. 2011). In the vine Mucuna, the flowers contain a small
concave mirror produced by two petals that also sends out signatures that enable
bats to distinguish between flowers with abundant nectar and those without (von
Helversen and von Helversen 2003).
The need for precision of pollen delivery, coupled with the ability of plants to
exploit the whole panoply of animal senses, leads to one logical question: how
specialized can these interactions get? For animals, as noted above, they must be
flexible and willing to exploit any resources that are available. As for plants, they
very seldom rely on one, two, or three species. However, there are a couple of
remarkably tight associations between specific groups of plants and pollinators.
Figures have been an evolutionarily active genus, with hundreds of species in
tropical and subtropical regions of several continents, and they rely on small
wasps for their pollination. Often, this reliance is so tight that one or a few species
of wasps pollinate a single fig species. The other genus that depends on very specific
pollinators is Yucca, which is pollinated by small moths. Yuccas and figs are often
cited as unusual examples of close specialization. However, detailed studies of both
associations indicate that, under specific circumstances, the plants can evolve novel
solutions to their needs for pollinators (Patel et al. 1993; Dodd and Linhart 1994).
There are many unresolved questions about the influences of floral signals upon
pollinators; they await the next generation of students of pollination. One of the
more contentious questions has focused on why bird-pollinated flowers tend to be
red. It seems that at least in hummingbirds, there is no innate attraction to the color
red, and it has been suggested that that bird flowers are red because they are not as
easily detectable by various bees, which means there is less competition for nectar.
However, it is not that simple. For starters, not all reds are created equal: reds
vary in their wave lengths, and then some reflect ultraviolet (UV) light, others
absorb it. The UV reflectors may attract bees, while the UV absorbers do not.
In addition, different groups of pollinating birds have somewhat different
visual systems. Recent work provides more details. For example, in a detailed
analysis of 206 plant species in Australia, Shrestha et al. (2013) demonstrated
that bird-pollinated and insect-pollinated flowers differ significantly in the chromatic cues of their flowers and that, although there is a good deal of variation
among the species, the wavelengths involved are concentrated near the optima
useful for discrimination by the two groups.
96
Y. Linhart
97
These styles can be either long or short, a condition known as distyly. Given
individuals produce only flowers of one type or the other. In order to effect pollination
only pollen from the opposite style length will be acceptable to a given individual.
The most famous distylous species are primroses. Darwin became intrigued by this
phenomenon, and his studies of primroses contributed significantly to his ideas about
evolution. As the horticulturist Henry Mitchell once pointed out, who knew that from
those modest primroses one could develop such revolutionary concepts. Tristylous
species are rare but operate on a similar principle. Another form of separation of floral
parts involves both the location of male and female parts within the flower and the
timing of floral development. For example, in many species with tubular flowers, as
the flower opens, the first organs to be exposed are the anthers that carry pollen. After
some time, typically 13 days, the anthers dry up and the style elongates, so that the
stigma protrudes furthest out of the corolla and is most likely to receive pollen from
another flower. In some species, including mimulus and some members of the family
Bignoniaceae, the stigma has two lobes that can close and prevent pollen deposition
in response to touch. This has been posited to be an adaptation to prevent selfpollination, but the evidence is modest.
Physiological mechanisms also prevent self-pollination. These are called
self-incompatibility and basically involve the ability of a plant to differentiate
between self-produced pollen and pollen from another plant. The former
either cannot germinate or the pollen tube grows more slowly down the style.
Such self-incompatibility is very common throughout the plant kingdom.
When pollinators are unreliable, plants can also manipulate these systems in other
ways. For example, at the periphery of species distributions or on islands, populations
evolve away from these mechanisms. Thus, on islands, species characterized by distyly,
including the primroses noted above have evolved to become homostylous. Yuccas and
other species that are characteristically self-incompatible in the center of their ranges
evolve towards being self-compatible on the periphery (Dodd and Linhart 1994).
Once pollen has fertilized an ovule, some plants can still have some control over
the genetic quality of their offspring. For example, if they have two or more
embryos within a single seed (a condition known as polyembryony), there can be
competition among embryos, and the slower-growing ones fall by the way side, and
only one emerges at germination and feeds on the seed resources. This condition is
known in several grasses including corn, rye, and wheat and also in many conifers
such as pines and Araucaria.
About those self-pollinators, it is known from basic genetics that inbreeding is
deleterious, so what about those species that self-pollinate their flowers?
These species have several features that mitigate the consequences of such inbreeding. First, natural selection has reduced the frequency of deleterious alleles to such
an extent that the probabilities of homozygous combinations with lethal effects are
very low. When they do occur, the seeds are simply aborted, and this is no great
loss, as such species typically produce many hundreds to thousands of seeds per
reproductive episode. Second, many of these species are polyploid, which provides
a reservoir of genetic variability. Third, there is periodic outcrossing in many of
these species, which replenishes the reservoir of variability needed.
98
Y. Linhart
Some species have yet another reproductive mechanism that can be useful:
apomixis which involves the ability to reproduce without fertilization. This is
especially useful when pollinators are unreliable but seeds must be produced at
all costs. It is an especially useful attribute in many weeds. However, even in
species that are apomictic, some opportunity for pollination and the associated
recombination is often maintained. For example, the dandelion Taraxacum
officinale is a well-known apomict and notorious weed. Yet it still produces pollen
and nectar (not needed if seeds are simply produced via mitosis) and is visited by
many insects. Careful analyses show that indeed at least some seeds in some plants
are produced sexually and help maintain genetic variability (Richards 2003).
99
100
Percentage of flowers visited
80
60
40
20
20
60
100
Distance (meters)
140
100
Y. Linhart
mating is random. This effective population size is strongly influenced by pollination. As might be expected, Ne tends to be smaller in plants that are self-compatible
and that are pollinated by small insects for which energetic constraints and optimal
foraging limit long-distance movement. In contrast, larger pollinators, including the
trapliners discussed above, as well as wind pollination usually lead to the development of larger population sizes. Whenever plants occupy environments that are
ecologically heterogeneous, such as mosaics of soil conditions, strong elevational
gradients, variable moisture, or light, these living conditions impose natural selection, and the evolutionary response to this selection will depend on the extent of
gene flow. In other words, gene flow is a homogenizing force across landscapes
unless it is limited. This means that small Ne will promote genetic differentiation
across small distances.
101
morphs which are either sweet or skunky to the human nose. Sweet-smelling flowers
are pollinated by large bees and have wider corolla lobes, have longer corolla tubes,
and are generally more flared out to accommodate these large bees. Skunky flowers
are pollinated by small flies, whose body mass is about 1.4 % that of the big bees.
They are smaller but produce just as much nectar per flower. Populations of
Polemonium often have both morphs, but bees provide about 75 % of the visits in
treeless tundra but only about 10 % in the lower zone of scattered trees called
krummholz. Flies show the reverse pattern (Galen 1989).
Such intra-specific variation in plants can indeed be an agent of genetic differentiation, but how often it leads to speciation is still open to debate and in need of
further studies with a broad array of species. These issues and their complexities are
discussed in detail by Kay and Sargent (2009) who conclude that floral differences,
and the associated differences in pollinator identity and behavior, are rarely if ever
sufficient to lead to speciation by themselves, while Schiestl (2011) suggests that
they may.
102
Y. Linhart
that must be partitioned in some optimal way, the need to complete ones life cycle
before harsh conditions set in, and various demands associated with genetic variability (Galen 1999).
Dispersal
How Do Plants Benefit?
There are multiple reasons why it is advantageous for seeds to be transported away
from their seed parent and from each other. All of them can exert strong selection
pressures favoring dispersal.
Escape
Seeds need to move away so as to reduce the likelihood that they will be damaged
by herbivores, parasites, or disease organisms that befall their seed parent. In
addition, when seeds and the seedlings they produce are in high densities, such
settings increase the likelihood of density-dependent attacks by seed or seedling
consumers.
Improved Growing Conditions
Soils near adult plants may be depleted of nutrients and/or have less water, more
shade, and perhaps an accumulation of toxic secondary compounds such as terpenes
leached into the soil from mother plants.
Colonization of New Habitats
Whenever a plant can establish in a habitat where it was absent before, it may
benefit for a variety of reasons, including increases of population size and escape
from herbivores and other consumers and diseases.
Genetic Recombination
Variability is a basic requirement for survival and adaptive evolution. Given that
once established plants will most likely exchange genes with near neighbors, if such
neighbors are genetically related, inbreeding ensues, and the next generation will
suffer the consequences. Conversely, if neighbors are somewhat different, the next
generation can benefit from being more variable.
103
Seed Packaging
At their simplest, seeds are covered by a hard envelope that protects them from
fluctuations in temperature and moisture. If they have no structural modifications,
they tend to be round or ovoid. Such seeds will simply fall to the ground when the
structure within which they develop matures and crack as it dries up. Obviously if they
fall, they have not traveled far from the mother plant or from each other, and, as noted
above, this is often problematic. There are situations when such limited dispersal is
useful, and they often involve plants adapted to live in very specific environments. For
example, plants living in temporary pools surrounded by dry habitats restrict their
dispersal to those pools. Plants that live on islands have often evolved reduced
dispersal abilities because it does not pay to get dispersed into salt water. However,
in general, there has been strong selection favoring devices that help the seeds travel
away from mother plants. Solutions to the challenge of dispersal come in many
shapes. These include having some way to exploit wind or water currents.
To get dispersed by animals, seeds must either attach themselves to a disperser
or offer a reward. Those that attach themselves tend to do so with hooks, bristles,
and barbs or have adhesive surfaces. A look at ones socks after a walk through a
dense grassland or a weed patch illustrates the effectiveness of such mechanisms.
Dwarf mistletoes of the genus Arceuthobium employ a different method. Their
seeds are inside fruit that at maturity are very sensitive to touch. When touched,
they explode and send sticky seeds out at a speed approaching 100 km/h. These
seeds then travel along, attached to the visiting bird or mammal until they are
rubbed off.
Rewards come in two major categories. The most common ones are in the form
of fleshy fruits, which often have bright, visually attractive coloration. In some
plants, the seeds alone are large enough to be attractive to animals that collect them,
transport them to specific locations and cache them for future consumption. Examples of such large seeds include the oaks, chestnuts, walnuts, hazelnuts, pistachios,
pinon pines, and their relatives (Fenner and Thomson 2005).
Wind
Wings attached to the seeds offer one solution and are common in groups as diverse
as grasses, conifers, ashes, and maples. Other devices that exploit wind are parachutes such as the ones found in dandelions or plumes of various shapes and many
other daisy relatives. Seeds can also be so small that they behave like dust particles
and are dispersed by the slightest breeze. Orchids use that solution.
104
Y. Linhart
Water
Coconuts provide a perfect example of the benefits of water for long-distance
dispersal: they are buoyant and protect their seeds from seawater well enough to
stay viable for many months. As a result they can colonize shores far distant from
their place of origin. They are pantropical in distribution, and it has been suggested
that they achieved this long-distance travel entirely on their own. Of course they are
also eminently edible, so human-aided movement on boats may have also been
important. It is sure that no one can agree on their geographical origin.
Water-mediated dispersal is uncommon and poorly studied. Species that rely on
it are few, with the notable exception of trees and shrubs called mangroves, which
belong to some 20 different families and have all adapted to life in coastal wetlands.
As such they are very important, for they provide the structural frameworks for
coastal ecosystems in the tropics and subtropics. In most mangroves, the seeds
germinate while still attached to the plants, so that the units that are dispersed are
actually seedlings (Kathiresan and Bingham 2001).
Ants
They are the only insects involved on a regular basis as seed dispersers, and they
play important roles in some specific settings. In temperate habitats of the Northern
hemisphere, some 300 plant species depend primarily on ants. They tend to be herbs
of the forest floor such as anemones, cyclamens, trilliums, and violets. In contrast,
in Australia and South Africa, the plants tend to be shrubs inhabiting dry
sclerophyllous fire-prone woodland. To attract the ants, the plants produce food
bodies called elaiosomes that are attached to the seeds. Ants transport these items to
their nests, where they eat the elaiosomes and discard the seeds. These seeds fall on
refuse piles, which provide more nutrients than surrounding soils. The germinating
seedlings thus get an extra boost in their early life (Gomez and Espadaler 2013).
Vertebrates
The species involved in seed dispersal are a remarkably diverse array of vertebrate
groups. Birds, rodents, and bats are the most frequent contributors. At this time, it is
impossible to ascertain the exact proportion of species in these various groups that
are involved as seed dispersers. However, it is estimated that over 1/3 of terrestrial
bird species eat fruit and about 1/5 of terrestrial mammals do so. Some important
seed dispersers are unexpected as they include various carnivores such as maned
wolves, coyotes, foxes, jackals, and even tigers. Figure 5 illustrates the diversity of
fruit and seed shapes and sizes seen at just one location in a tropical forest in
Ecuador. These fruits and seeds will be dispersed by many species of birds and
mammals.
In addition, in riparian habitats especially in the tropics, but in other regions as
well, fishes are important fruit dispersers and have been so for a very long time,
perhaps since the Paleozoic. Indeed, it may be that they were the first vertebrates to
act as seed dispersers. At least 275 species are frugivorous. They belong to various
groups including piranhas, catfishes, carps, and minnows. They are especially
important in Neotropical forests that are periodically flooded such as the Brazilian
105
Fig. 5 Diversity of seed and fruit types in a forest in Ecuador. The plant genera represented
include Spondius sp. (Anacardiaceae, mango and sumac family), Schefflera (Araliaceae, ivy and
ginseng family), Raffia (Arecaceae, palms), and Guarea (Meliaceae, mahogany family). Other
families include the Annonaceae (sweetsop family), Araceae (anthurium Fig. 1 family), and
Lauraceae (laurel family) (Photo and information courtesy of K.M. Holbrook)
Pantanal where they help disperse a large number of shrubs and trees and can be the
principal dispersers for many trees (Galetti and Goulding 2011). Other inhabitants
of riparian regions can also be useful to plants in that regard. For example, among
the Crocodylia, at least 13 species are documented as seed dispersers, and
they consume seeds or fruits in at least 46 genera belonging to 34 families
(Platt et al. 2013). Lizards and turtles are also known to get involved in seed
dispersal, especially on islands (Olesen and Valedo 2003), and one species of
frog has been reported as a frugivore so far (da Silva et al. 1989).
Fruit Characteristics
Visual
Ripe fruits dispersed by birds are often brightly colored, and the predominant colors
are red and black, and often reflect ultraviolet light. In contrast, mammal-dispersed
fruit are not as visually striking and tend to have more subdued colors. The fact that
colors can promote specificity of dispersers is illustrated by the bright red color of
ripe chili peppers. The red color is produced by capsaicins which attract the birds
106
Y. Linhart
and have no effects on their palates. Conversely, they produce a memorably spicy
burning sensation in mammals. This helps keep mammalian frugivores away from
the fruit (Schulze and Spiteller 2009).
Olfactory
Mammals have more highly developed olfactory abilities than birds, so detectable
smells are often important features of mammal-dispersed fruits. Bat-dispersed fruit
including various figs are best known in that context. This area of plant-disperser
interactions is poorly understood and a wide-open field for study. It is sure that
certain fruits such as durians which live in the forests of Malaysia and Indonesia are
very good advertisers in that context. Their taste is heavenly, but the smell they emit
has been described as fermenting dirty socks with elements of onions and leaking
gas and a background hint of long-dead corpse. Given the wonderful complexity of
this bouquet, it is not surprising that they attract diverse local denizens including
tigers, elephants, and monkeys that disperse them with gusto. See, for example, a
video of a tiger checking out a durian (Sumatran tiger inspects durian fruit on forest
floor www.arkive.org Species Mammals Tiger). As for ginkgo, there is no
idea what animals might have dispersed its stinky fruit. Unfortunately, it is now a
living fossil that grows only in urban habitats. It has disappeared from the wilds of
Asia but has left enough fossils around so it is well known that it was a forest
dweller about 200 million years ago. Whether it was dispersed by dinosaurs or
small mammals, or both, will never be known.
Size and Nutritional Value
There is a large range of sizes in fruit, and there are very general patterns of
variation between geographic regions. Thus, in multiple families, fruits are on
average larger in the tropics of Asia and Africa than in the Americas, presumably
because of the absence of larger seed dispersers in the latter regions. As for fruit
composition, there is also a large range of variation of fruit composition: some (e.g.,
think cherries, apples, or citrus) are mostly carbohydrates and offer little reward to
the dispersers beyond a quick energy boost, while others are rich in proteins (e.g.,
avocado, guava, dates) and lipids (e.g., olives, as well as magnolia, dogwoods
(Cornus spp.), and Virginia creeper (Parthenocissus)) (Johnson et al. 1985).
Patterns of Dispersal
The diversity of seed dispersal mechanisms is clearly the outcome of
strong evolutionary pressures generated by the advantages listed above. However,
the behavior of frugivores following ingestion is highly variable and often
poorly known: as a result, the shapes of dispersal distributions away from sources
are erratic and usually not quantified. In addition, the diversity of dispersers
in natural landscapes, especially in the tropics, is very high, so studies that focus
on just a few species cannot provide an accurate picture of dispersal patterns
(Table 1).
Table 1 Routine
maximum seed dispersal
distances achievable by
various combinations of
plants and dispersal agents
Distance
010 m
10100 m
100 m1 km
110 km
>10 km
107
Vector (propagule type)
Mechanical
Ants
Wind (large winged fruits)
Rodents
Macaques (large seeds, not swallowed)
Small- and medium-sized forest passerines
Fruit bats (large seeds)
Most primates (seeds swallowed)
Large canopy birds
Open-country passerines
Small fruit bats (tiny seeds)
Orangutans
Carnivores, including civets, martens, and bears
Most terrestrial herbivores
Wind (tiny seeds), water
Fruit pigeons
Large fruit bats (tiny seeds)
Elephants, rhinoceroses
People
For this reason, until recently, the only species for which reasonable data on
dispersal patterns was available were for wind-dispersed seeds collected from containers at varying distances from sources. For such species, the patterns are straightforward: the distributions tend to be leptokurtic, and small numbers of seeds
can travel several km. For animal-dispersed species, distributions definitely show
attenuation with distance, but over those distances, they are often very patchy (Cain
et al. 2000). However, the use of DNA microsatellite analysis is documenting the
complexities of such dispersal very nicely. For example, Jordano et al. (2007) have
found that both birds and mammals disperse the fruit of Prunus mahaleb; small birds
tend to move seeds shorter distances and into covered habitats, while mammals move
them longer distances and into open areas and also account for about two-thirds of
introduction of immigrant seeds into populations (Fig. 6). Other studies document
long-distance patterns and show that the tails of distributions can be much longer up
to several km than previously thought (Ashley 2010).
One important question is the extent to which the dispersers actually deliver the
seeds to locations where the seeds can get established. These are often known as
safe sites, and such dispersal to useful locations is usually called directed dispersal
(Wenny et al. 2011). Such dispersal is becoming recognized as an important
alternative to the notion that all seeds are dispersed as clouds over the landscape.
The best understood examples in order of the number of plant species involved
include (1) ant-dispersed plants with seeds attached to elaiosomes. The ants transport the seeds to their nest areas. At least 3000 species in 60 plant families have
108
Y. Linhart
200
175
150
100
Small birds
Thrushes
Corvids/Pigeons
Carnivorous mammals
75
125
50
100
75
25
50
0-50
100
150
200
250
300
350
400
450
500
550
600
650
700
750
800
850
900
950
25
Distance (m)
0
Immigrant
seeds
Fig. 6 Frequency distribution of seeds of the cherry Prunus mahaleb dispersed by small birds
including warblers Sylvia spp., and robins Erithacus rubecula; thrushes (Turdus spp.); large birds,
including pigeons Columba, and corvids such as carrion crows Corvus corone; and carnivorous
mammals including red fox Vulpes vulpes, marten Martes foina, and badger Meles meles (Adapted
from Jordano et al. 2007)
109
110
Y. Linhart
One solution that plants have evolved is to protect these valuable packages with
various toxic compounds (see Chap. 6, Evolutionary Ecology of Chemically
Mediated Plant-Insect Interaction). Another is to put out very large numbers of
seeds simultaneously so as to overwhelm the seed consumers, but do so on an
irregular basis so that the consumers cannot track those bonanzas. This phenomenon
is known as masting, and for any one species in any one location, masting episodes
occur every few years. Many species of trees and other perennials follow this pattern,
which also has the advantage of synchronizing flowering thus improving the probability of pollination and outcrossing for all members of the participating population
(Kelly and Sork 2002).
111
112
Y. Linhart
113
The issue is that these weeds can and do exchange genes with their cultivated
relatives. It must be borne in mind that in todays agriculture, many crops have been
bred to be herbicide-resistant themselves. The rationale is that if crop plants are
herbicide resistant, then herbicides can be used in crop fields with impunity to
control the weeds, which is much cheaper than other means of weeding. Beautiful
logic, until biology intervenes. The problem was predicted by the work of the
botanist Jack Harlan who was the first to draw attention to the fact that crop plants
often grew in close proximity to weeds that were close relatives. For example, he
observed that in Mexico and Central America, maize grew in the company of its
ancestor and competitor, teosinte. In Africa, he saw cultivated and weedy sorghum
in close association, in Asia cultivated and weedy rice grew side by side, and so
on. These observations coupled with the recognition that these weeds and crops
could interbreed led him to formulate the concept of the compilospecies which
posited that whenever groups of species were closely related, they could exchange
genes and thereby compile useful information. In retrospect, it is no wonder that
herbicide resistance has evolved so quickly in so many species. We have helped the
process along: we have introduced genes for herbicide resistance into crops, whose
pollen and seeds move about, sometimes great distances (as per Ashley 2010), and
help pass on those genes to weedy relatives.
As for the escape of transgenes, this possibility was brought up at least two
decades ago. So far, it seems that relatively few transgenes have ended up in wild
populations, but still, thanks to unexpected dispersal of pollen and/or seeds, they are
found in settings where they were not intended to be (Ellstrand 2012). On issue is that
escape into wild populations is not the only problem. Escape of multiple transgenes
into populations free of such genes is another. This is happening in Mexico. This is
very problematic given the dependence upon maize as a food crop in humans
worldwide and because Mexico is the original home of maize and the center of
diversity of this species; at least 60 distinct land races adapted to very different
ecological conditions, and several wild relatives of maize are unique to the country.
This genetic diversity represents a very important reservoir for future breeding of
maize. The majority of maize fields in Mexico are small, family enterprises, and
seeds are usually replanted within the area where they were produced. This method
contributes to the maintenance of these land races. The accidental introduction of
foreign transgenes into such varieties can disrupt the integrated nature of their
genomes. If one imagines that the genome of a variety is like a blueprint that guides
its construction, the sudden introduction of new components into the design can alter
the appearance and/or function of the finished product. In addition, the blueprints are
no longer useful for future work. For these reasons, there was concern about the
introduction of genetically engineered corn in Mexico, and a moratorium on such
introduction was put in place in 1998. Despite this moratorium, transgenes have been
detected in native populations, and the consequences of these careless introductions
are being assessed (Pineyro-Nelson et al. 2009).
As for transgenes for herbicide resistance, they are also spreading in our landscapes and creating problems as illustrated in this case study. Creeping bent grass
(Agrostis stolonifera) is commonly used in golf courses. In 2002, a version of this
114
Y. Linhart
species carrying genes for resistance to the herbicide glyphosate (aka Roundup )
was planted by the Scotts Company on 162 ha in Oregon. Wind-dispersed pollen
carrying the resistance genes moved from that population and fertilized ovules of
two local species (A. stolonifera and A. gigantea), and the hybridizations occurred
on sentinel plants as far as 21 km away. In addition, winds helped move transgenic
seeds into nearby areas. Recently the situation has become more complicated
because of the detection of an intergeneric hybrid which carries the transgenes
and consists of a combination of the bent grass with rabbit-foot grass (Polypogon
monspeliensis) (Snow 2012).
Overall, given the warning provided by Ashley (2010) about the fact that longdistance gene flow via pollen and seed is much more prevalent than we thought, the
message is clear. . .there are problems afoot.
Future Directions
There are over 250,000 species of flowering plants in the world (8090 % are
pollinated by animals) and another 1,000 or so species of non-flowering plants that
disperse by seed. There are probably well over 30,000 species of animals involved
in the tasks of pollination or seed dispersal. It is no wonder that we still have much
to learn about the interactions. The issues that are especially poorly known are
noted in the text and summarized below:
The variability of the color spectra and UV nectar guides produced by plants to
attract animals and the ability of various animals to detect those signals. In more
general terms, the intricacies of visual and chemical communication in the
contexts of pollination, seed dispersal, and herbivory deserve greater attention
(Schaefer and Ruxton 2011).
The extent to which shifts in signals, especially olfactory and visual ones, can
produce shifts in pollinator visitation patterns and resulting gene flow and
population differentiation is also open to question.
Animal-mediated seed dispersal outside of the temperate zones of North
America and Europe is a very open field, both in the tropics and in the Southern
Hemisphere. Even within temperate areas, we still have a very limited understanding of the ecosystem services that birds provide (Wenny et al. 2011).
Large vertebrate dispersers are major contributors to seed dispersal networks,
especially in the tropics (Table 1), but much of our information is anecdotal.
Their real contributions are poorly known because they are difficult to study and
can be very rare. In addition, some are already missing from some ecosystems
(Corlett 2009; Vidal et al. 2013).
This introduction to pollination and dispersal should be used in combination
with the discussion of herbivory (Chap. 6, Evolutionary Ecology of
Chemically Mediated Plant-Insect Interaction) and biodiversity and population
dynamics (Chap. 2, Plant Biodiversity and Population Dynamics) to get a
synthetic understanding of linkages among populations, metapopulation
dynamics in space and time, and long-term dynamics.
115
Summary
Plants are stationary and depend on external agencies to help them reproduce and
disperse their seeds. Most plant species utilize animal pollinators and seed
dispersers, although in specific ecosystems some plants can use wind or water for
such transport.
To attract these animal vectors, plants use various food rewards including pollen,
nectar, seeds, and fruits.
In terms of species numbers, the majority of pollinators are insects, and the
majority of seed dispersers are vertebrates.
The genetic structure of plant populations is strongly influenced by their pollen
and seed vectors. When wind is the dispersing agent, pollen and seed movement are
relatively straightforward and can be described by leptokurtic distributions, with
most of the pollen grains or seeds transported short distances and tails extending
long distances away from the source. When the dispersal is by animals, the
behaviors of individuals and species are so variable as to render generalizations
difficult.
The ecology and evolution of plants is not just about plants: animals are
important actors in those plays. Consequently, the effects of humans upon pollinators and seed dispersers should influence management decisions in natural and
agronomic ecosystems.
References
Adler LS, Irwin RE. What you smell is more important than what you see? Natural selection on
floral scent. New Phytol. 2012;195:5101.
Ashley MV. Plant parentage, pollination, and dispersal: how DNA microsatellites have altered the
landscape. Crit Rev Plant Sci. 2010;29:14861.
Barrett SCH. The evolution of plant sexual diversity. Nat Rev Genet. 2002;3:27484.
Buchmann SL, Nabhan GP. The forgotten pollinators. Washington D.C.: Island Press; 1996.
Cain ML, Milligan BG, Strand AE. Long-distance seed dispersal in plant populations. Am J Bot.
2000;87:121727.
Cheptou PO, Carrue O, Rouifed S, Cantarel A. Rapid evolution of seed dispersal in an urban
environment in the weed Crepis sancta. Proc Natl Acad Sci. 2008;105:37969.
Cody ML, Overton JM. Short-term evolution of reduced dispersal in island plant populations.
J Ecol. 1996;84:5361.
Corlett RT. Seed dispersal distances and plant migration potential in tropical East Asia. Biotropica.
2009;41:5928.
da Silva HR, de Britto-Pereira MC, Caramaschi U. Frugivory and seed dispersal by Hyla truncata,
a neotropical treefrog. Copeia. 1989;7813.
Delye C, Jasieniuk M, Le Corre V. Deciphering the evolution of herbicide resistance in weeds.
Trends Genet. 2013. In press.
Dodd RJ, Linhart YB. Reproductive consequences of interactions between Yucca glauca
(Agavaceae) and Tegeticula yuccasella (Lepidoptera) in Colorado. Am J Bot. 1994;81:81525.
Ellis AG, Johnson SD. Floral mimicry enhances pollen export: the evolution of pollination by
sexual deceit outside of the Orchidaceae. Am Nat. 2010;176:E14351.
Ellstrand NC. Over a decade of crop transgenes out-of-place. In: Regulation of agricultural
biotechnology: the United States and Canada. The Netherlands: Springer; 2012. p. 12335.
116
Y. Linhart
Fenner M, Thomson K. The ecology of seeds. New York: Cambridge University press; 2005.
Fenster CB, Armbruster WS, Wilson P, Dudash MR, Thomson JD. Pollination syndromes and
floral specialization. Annu Rev Ecol Evol Syst. 2004;35:375403.
Galen C. Measuring pollinator-mediated selection on morphometric floral traits: bumblebees and
the alpine sky pilot, Polemonium viscosum. Evolution. 1989;43:88290.
Galen C. Why do flowers vary? BioScience. 1999;49:63140.
Galetti M, Goulding M. Seed dispersal by fishes in tropical and temperate fresh waters: the
growing evidence. Acta Oecol. 2011;37:56177.
Gaskett AC. Orchid pollination by sexual deception: pollinator perspectives. Biol Rev.
2011;86:3375.
Gomez C, Espadaler X. An update of the world survey of myrmecochorous dispersal distances.
Ecography. 2013;36:11931201.
Gomez JM, Perfectti F, Bosch J, Camacho JPM. A geographic selection mosaic in a generalized
plantpollinatorherbivore system. Ecol Monogr. 2009;79:24563.
Hamrick JL, Trapnell DW. Using population genetic analyses to understand seed dispersal
patterns. Acta Oecol. 2011;37:6419.
Hu S, Dilcher DL, Jarzen DM, Winship D. Early steps of angiospermpollinator coevolution. Proc
Natl Acad Sci. 2008;105:2405.
Irwin RE, Strauss SY, Storz S, Emerson A, Guibert G. The role of herbivores in the maintenance of
a flower color polymorphism in wild radish. Ecology. 2003;84:173343.
Johnson RA, Willson M, Thomson JN, Bertin RI. Nutritional values of wild fruit and consumption
by migrant frugivorous birds. Ecology. 1985;66:81927.
Jordano P, Garcia C, Godoy JA, Garca-Castano JL. Differential contribution of frugivores to
complex seed dispersal patterns. Proc Natl Acad Sci. 2007;104:327882.
Karron JD, Ivey CT, Mitchell RJ, Whitehead MR, Peakall R, Case AL. Viewpoint: part of a special
issue on plant mating systems. Ann Bot. 2012;109:493503.
Kathiresan K, Bingham BL. Biology of mangroves and mangrove ecosystems. Adv Mar Biol.
2001;40:81251.
Kay KM, Sargent RD. The role of animal pollination in plant speciation: integrating ecology,
geography, and genetics. Annu Rev Ecol Evol Syst. 2009;40:63756.
Kearns CA, Inouye DW. Techniques for pollination biologists. Niwot: University Press of
Colorado; 1993.
Kelly D, Sork VL. Mast seeding in perennial plants: why, how, where? Annu Rev Ecol Syst.
2002;33:42747.
Kessler D, Diezel C, Clark DG, Colquhoun TA, Baldwin IT. Petunia flowers solve the defence/
apparency dilemma of pollinator attraction by deploying complex floral blends. Ecol Lett.
2013;16:299306.
Levine JM, Murrell DJ. The community-level consequences of seed dispersal patterns. Annu Rev
Eco Syst. 2003;34:54974.
Linhart YB. Ecological and behavioral determinants of pollen dispersal in hummingbirdpollinated Heliconia. Am Nat. 1973;107:51123.
Llandres AL, Gawryszewski FM, Heiling AM, Herberstein ME. The effect of colour variation in
predators on the behaviour of pollinators: Australian crab spiders and native bees. Ecol
Entomol. 2011;36:7281.
Olesen JM, Valedo A. Lizards as pollinators and seed dispersers: an island phenomenon. Trends
Ecol Evol. 2003;18:17781.
Ollerton J, Winfree R, Tarrant S. How many flowering plants are pollinated by animals? Oikos.
2011;120:3216.
Patel A, Hossaert-Mckey M, Mckey D. Ficus-pollinator research in India: past, present and future.
Curr Sci. 1993;65:24353.
Patiny S. Evolution of plant-pollinator relationships. Cambridge: Cambridge University Press;
2011.
117
Contents
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Phenotypic Plasticity Is a Particular Form of Variable Phenotypic Expression . . . . . . . . . . . . . .
Definition of Phenotypic Plasticity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Phenotypic Plasticity Is Not the Only Mechanism that Generates Variable
Phenotypic Expression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Challenges in Defining Phenotypic Plasticity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
The Particular Importance of Phenotypic Plasticity in Plants . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Phenotypic Plasticity Is Often, but Not Always, Interpreted as an Adaptive
Response to Variable Environments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Phenotypic Plasticity as Adaptive . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Phenotypic Plasticity: A Highly Selected Trait or a Consequence of
Selection for Multiple Phenotypes? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Theoretical Limits to Selection for Phenotypic Plasticity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Phenotypic Plasticity as Nonadaptive or Maladaptive . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
The Role of Phenotypic Plasticity in Evolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Phenotypic Plasticity Versus Developmentally Programmed Changes in Phenotypic
Expression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Techniques for Evaluating Phenotypic Plasticity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Norms of Reaction Characterize Phenotypic Expression for One or More
Genotypes Across a Range of Environments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Use of Developmentally Sensitive (Common Size or Developmental Stage)
Comparisons Versus Common Time or Age Comparisons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Growth Analysis and Allometric Approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Modular Growth as a Platform for Evaluating Phenotypic Plasticity in Plants . . . . . . . . . . .
Selecting Methodological Approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Future Directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
120
121
121
121
122
123
124
124
124
125
126
126
127
127
127
128
133
134
136
138
138
119
120
Abstract
Phenotypic expression is the result of a complex interplay between an organisms genes and its environment.
During growth and development, organisms undergo a programmed series of
phenotypic changes. Phenotypic expression thus varies throughout growth
and development, even when the environment in homogenous and static. This
has been termed ontogenetic drift.
Phenotypic expression may also vary with environmental conditions. The
ability to vary phenotypic expression in response to environmental conditions
is known as phenotypic plasticity.
The ability of an organism to express variable phenotypes in heterogeneous
environments has been thought to confer adaptive benefits that increase
fitness. Plants, as immobile organisms, cannot relocate to more favorable
environments; plant phenotypic plasticity could be under strong selective
pressure in predictably variable environments.
Plant growth rates and developmental trajectories are generally plastic; i.e.,
they frequently vary with local environmental conditions.
Whenever environmentally induced plasticity in growth and development
occurs, interpretations of phenotypic plasticity are confounded with changes
in phenotypic expression associated with ontogenetic drift.
Plant phenotypic plasticity should be evaluated in a developmentally explicit
context. Phenotypic expression should be characterized in light of developmental trajectories of phenotypic change whenever possible.
Comparing plant phenotypes at a common age versus a common developmental stage may result in incorrect conclusions regarding the nature of the
observed phenotypic variation.
Selection of methodological approaches to evaluate plant phenotypic expression should align with the hypothesis under investigation.
Introduction
Biologists have developed a small handful of unifying themes to explain the
astonishing diversity of form and function exhibited by organisms. Phenotypic
plasticity is one of those themes that continues to fascinate biologists from
diverse backgrounds from ecologists and geneticists to developmental and evolutionary biologists. It is often a subject that students have difficulty grasping, for
phenotypic plasticity is the result of the interplay between two distinct but
interacting identities the genetics of an organism and its environment but is
responsible for much of the intraspecific variation observed in ecological contexts.
In this chapter, we will describe phenotypic plasticity, offer examples of how it can
confer putative adaptive advantages for species in predictably variable environments, explore how phenotypic expression is facilitated and constrained by
predetermined patterns of phenotypic expression throughout growth and development, and discuss methodological approaches to assessing phenotypic plasticity.
121
122
stochasticity, such that gene and protein expression, and thus trait development, can
vary at least in part due to small accumulations of chance events (e.g., Yampolsky
and Scheiner 1994).
Phenotypic variation that does not correlate with a specific genotype or specific
environmental cue, but is the result of stochasticity in the biochemical processes
involved in gene and protein expression and other cellular noise that occurs
throughout development, is referred to as developmental noise (Bradshaw
1965). If developmental noise generates sufficient variation in phenotypic expression, genetically identical individuals grown in the same environment will exhibit
different phenotypes. The resulting phenotype could be adaptive, maladaptive, or
neutral depending on the environmental conditions (DeWitt and Scheiner 2004).
For example, in times of stabilizing selection where a mean phenotype is more
desirable, developmental noise might reduce fitness; alternatively, variable phenotypic expression may increase the probability that at least some members of a
population are able to survive and reproduce in stressful or rapidly changing
environments.
123
Finally, Weiner (2004) argues that the definition of trait, the aspect of phenotype under evaluation as plastic or not, is itself too broadly defined and suggests that
the fact that a given trait is measurable does not guarantee that the trait is relevant to
the organisms ecological persistence or evolutionary success.
In light of the difficulties in developing a clear and consistent definition of phenotypic plasticity, one should take care to understand the context in which researchers
frame their individual questions. DeWitt and Scheiner (2004, p. 2) eloquently state:
Such breadth of scope reinforces the idea that a particular trait value as observed in a given
environment always is a special case of a potentially more complex relationship. That is,
specific phenotype-environment observations are a fraction of a multidimensional space.
This view promotes in our thinking the constant and useful caveat that given phenotype
distributions may only apply for the environment in which observation is conducted.
Extrapolation beyond given conditions must be justified rather than assumed.
124
125
126
127
128
129
(Coleman et al. 1994). Figures 1 and 2 depict two common patterns exhibited by
plants in variable environments: plastic growth responses (Fig. 1) and developmentally programmed changes in phenotypic expression or ontogenetic drift (Fig. 2).
If one were to compare phenotypes from these two environments at a common time,
they would correctly conclude that phenotypic expression differs in the two
environments, which may be interpreted as support for phenotypic plasticity.
However, in Fig. 3, the pattern of change in phenotypic expression throughout
growth and development is fixed (i.e., the slope is constant for each line).
The variable phenotypic expression is not apparent when plants are compared at
similar stages in growth and development. In other words, environmental heterogeneity causes plasticity in growth rates, and ontogenetic drift occurs, but the ontogenetic program of phenotypic change throughout development is constant (i.e., not
plastic). Thus, when compared at a common point in growth and development, there
are no phenotypic differences among treatments, although phenotypes will
differ when compared at a common time (Coleman et al. 1994). This trait is said
to exhibit apparent plasticity, which is defined as variation in a trait because
of environmentally induced variation in growth or development coupled with
130
Fig. 3 The phenotypic trait values from Fig. 2 are replotted against the growth values from Fig. 1,
using the time point data for each x,y pair. In this example, the ontogenetic trajectory for the
phenotypic trait does not vary with environment, that is, to say it is not plastic. Thus, variations in
phenotypic expression are consequences of plasticity in growth and developmental rates coupled
with ontogenetic drift and not plasticity in the ontogenetic program for phenotypic expression
ontogenetic drift in the trait of interest (McConnaughay and Coleman 1999; Wright
and McConnaughay 2002).
Apparent plasticity is usually the result of environmental conditions that the
plant does not have appreciable control over (toxins, soil nutrients, oxygen levels,
or temperature) as opposed to a condition the plant can respond to by actively
changing development in such a way as to maintain constant growth rates (Scheiner
1999). These environmental conditions alter biochemical processes within the
plant, which in turn affect development, typically resulting in smaller size plants
compared to their non-limited cohorts (Whitman and Agrawal 2009).
Those studying optimal partitioning theory (OPT) models will find this scenario
particularly relevant. For example, plants grown in shade conditions invest a greater
proportion of assimilated resources to growing leaves and will have a greater leaf
area ratio (LAR) than the same species grown in more abundant light conditions.
However, shade-grown plants (or plants in any unfavorable condition) typically
grow and develop more slowly, and plants typically invest more biomass in
structural support relative to leaf area as they grow. Is the reduced LAR in
shade-grown plants the result of structural and functional adjustments of resource
allocation as predicted by OPT or a consequence of slow growth rates and delayed
development along a fixed ontogenetic trajectory? If examined at a common age,
the conclusion may be in favor of OPT in which the plant is apparently optimizing
function, but if plants are compared at the same size, differences in allocation may
disappear, diminishing the discussion on the effects of differing light treatment on
LAR (Coleman et al. 1994).
Mooney et al. (1988) studied the effects of sulfur dioxide (SO2) on the growth
and resource acquisition of cultivated radish, Raphanus sativus, by measuring
changes in photosynthetic activity, biomass accumulation, and root to leaf allocation relative to controls. Excess atmospheric sulfur dioxide is caused by human
131
132
A third scenario is predicted in which the direction of the results is reversed, with
one treatment having a greater value, when compared at plant age, and a smaller
value compared to the other treatment, when compared at a common plant size or
vice versa (Coleman et al. 1994). Evans (1972) in his work with Impatiens
parviflora demonstrated this pattern in regard to leaf weight as a function varying
light intensity.
133
levels of light, nutrients, and water. This approach can be likened to the split-brood
design explained by Via (1993), in which a family member or clone is split among
different environments. If one tests random samples from a population, only mean
plasticity can be estimated and values for genetic variation in plasticity cannot be
obtained, this approach is probably more realistic as not all plants reproduce
asexually, but small differences in treatment may be confounded due to lack of
genetic similarity. In this experiment, when one resource was kept low, the other
two were maintained at moderate or high levels. Growth parameters were measured
over a period of 81 days with frequent harvests over time. Root allocation was
examined over time and as a function of total plant biomass (i.e., same ontogenetic
stage). When examined in both manners, root allocation did increase over the
81-day growth period in a direction that was opposite of that predicted of annual
herbs. When compared across time, all three treatments resulted in significant
differences between high and low resource levels, with an increase in allocation
to roots for low nutrient and low water treatments and an increase in allocation to
shoots in the light-limited treatments. When compared across size, water and
nutrient treatments remained significant; however, there was no longer a difference
between low- and high-light conditions. Therefore, root allocation in response to
light limitation resulted in slowed growth rates along a fixed ontogenetic trajectory
a response in agreement with apparent plasticity while root allocation patterns
exhibited in response to belowground resources (water and nutrients) were consistent with complex plasticity. These results agreed with predictions of the developmentally explicit model (McConnaughay and Coleman 1999). As this study
demonstrates, evaluating phenotypic plasticity in the context of ontogeny is important as it allows a more complete understanding of an organisms ability to respond
to a heterogeneous environment.
134
135
nodes and branches, leaves, and distance between internodes. Vegetative tissues
exhibit indeterminate growth patterns and during ontogenetic development could
produce node after node forming repeating units called modules. Meristematic
tissue also allows for horizontal growth of plants which is termed clonal growth.
Each individual that arises is genetically identical to the original plant and is
referred to as a ramet of the larger genet; physical connections to exchange nutrients
may be temporary or long lasting (Silvertown and Charlesworth 2001). However,
SAMs can also exhibit determinate growth patterns if production of inflorescences
or flowers occurs.
136
primordium depends upon where and when it is produced within the architecture of
an organism and what events [environmental conditions] have preceded it during
ontogeny (Diggle 1994, p. 1354).
The benefit of modular integration is that it can help a plant overcome spatial and
temporal environmental variation that decreases their success. Traditional methods
of studying plasticity, such as developmental reactions norms, do not explicitly
recognize that the whole plant phenotype is the integrated sum of many modules
that may develop and exist in different environmental states and that plasticity may
be expressed at the level of modules (de Kroon et al. 2005).
137
herbivory, intra- and interspecific competition, plant-plant or plant-animal interactions, or analysis of plant-environment interactions, such as those related to changes
in season. Examples of phenotypic processes that may benefit from this approach
include leaf nitrogen levels in relation to herbivory, flowering loads and pollinator
activity, or total fitness in annual plants affected by end of season frosts (Coleman
et al. 1994; Wright and McConnaughay 2002). Agrawal et al. (2012) examined
plant resistance mechanisms in Oenothera biennis. By suppressing insect herbivores in the field, they found that after only 5 years, protected plants diverged from
control plants, producing less defensive chemicals in their fruits while increasing
competitive ability likely due to greater energy available for biomass allocation.
The approach of chronological age was appropriate for this real-time study that
demonstrated rapid, ecological, and evolutionary change.
The examination of two or more functionally related phenotypic traits is highly
dependent upon the degree of ontogenetic drift that each trait exhibits. An allometric approach is best used to assess each of these traits individually and in relation to
each other. Allometric growth is an unequal change in the size of one body part
relative to the change in size of another body part, or sometimes the entire body,
whereas isometric growth is the condition of directly proportional change among
body parts arising from identical growth rates of the individual parts (WestEberhard 2003; Wu et al. 2003). Most traits exhibit allometry with respect to one
another, as opposed to isometry (West-Eberhard 2003). One example of isometry in
plants is that for every leaf there will be exactly one petiole attaching the leaf to the
stem (though the sizes of the leaf blade and petiole in question may be allometric).
Isometry between two traits would be characterized by a simple linear relationship
with a slope of 1.0. Any deviation from a slope of 1.0 represents an allometric
relationship between the traits such that change in one trait during growth and
development is greater or lesser than the change in the other trait. If the relationship
between two traits exhibits curvilinearity during development, the traits are said to
exhibit complex rather than simple allometry (Coleman et al. 1994). When structures compete with each other for resources (like roots and shoots), a change in
either allometric ratio of either trait would be expected to influence the other (WestEberhard 2003). Other examples of functionally related phenotypic traits include
root to shoot biomass accumulation, comparisons of reproductive versus vegetative
biomass, relationships between height and diameter, tissue carbon to nitrogen
ratios, or leaf nitrogen composition and photosynthetic relationships (Coleman
et al. 1994; Wright and McConnaughay 2002).
Ontogenetic drift can only be ignored when relationships between biomass
variables are isometric and linear (i.e., simple allometry). When biomass allocation
patterns are allometric, patterns will differ throughout growth and development
regardless of environmental conditions, meaning they exhibit ontogenetic drift
(McConnaughay and Coleman 1999). Many allocation patterns follow allometric
trajectories, which are intrinsically a function of plant size (Weiner 2004), so any
factor that influences size will change allocation. It is well known that plant
allocation patterns are size dependent, but methods traditionally used to assess
biomass allocation, such as optimal partitioning theory, make the assumption that
138
Future Directions
Studies of plant phenotypic plasticity aim to increase our understanding of how
plants cope with variable environments. At this time, no simple unified theory exists
that predicts when, how, or to what extent plants can respond to changes in the
environment with changes in phenotype or under what circumstances any such
phenotypic changes will increase fitness. Past work has obscured our evaluation of
phenotypic plasticity by confounding environmentally induced variation in phenotypic expression with environmentally induced variation in growth and development and developmentally fixed patterns of phenotypic expression. A more
developmentally explicit approach to evaluating the mechanisms of variable phenotypic expression could lead to a greater understanding of the limits of phenotypic
plasticity and its potential significance in evolutionary and ecological contexts.
References
Agrawal AA, Hastings AP, Johnson MTJ, Maron JL, Salminen J. Insect herbivores drive real-time
ecological and evolutionary change in plant populations. Science. 2012;338:1136.
Bloom AJ, Chapin III FS, Mooney HA. Resource limitation in plants an economic analogy.
Annu Rev Ecol Evol Syst. 1985;16:36392.
Bradshaw AD. Evolutionary significance of phenotypic plasticity in plants. Adv Genet.
1965;13:11555.
Coleman JS, McConnaughay KDM. A non-functional interpretation of a classical optimalpartitioning example. Funct Ecol. 1995;9:951954.
Coleman JS, McConnaughay KDM, Ackerly DD. Interpreting phenotypic variation in plants.
Trends Ecol Evol. 1994;9:18791.
de Kroon H, Heidrun H, Stuefer JF, van Groenendael JM. A modular concept of phenotypic
plasticity in plants. New Phytol. 2005;166:7382.
DeWitt TJ, Scheiner SM. Phenotypic plasticity: functional and conceptual approaches. New York:
Oxford University Press; 2004.
DeWitt TJ, Sih A, Wilson DS. Costs and limits of phenotypic plasticity. Trends Ecol Evol.
1998;13:7781.
Diggle PK. The expression of andromonoecy in Solanum hirtum (Solanaceae): phenotypic plasticity and ontogenetic contingency. Am J Bot. 1994;81:135465.
Evans GC. The quantitative analysis of plant growth. Oxford: Blackwell Scientific; 1972.
Fitter A, Hay R. Environmental physiology of plants. 3rd ed. London: Academic; 2002.
Garland T, Kelly SA. Phenotypic plasticity and experimental evolution. J Exp Biol.
2006;209:234461.
Gedroc JJ, McConnaughay KDM, Coleman JS. Plasticity in root shoot partitioning: optimal,
ontogenetic, or both? Funct Ecol. 1996;10:4450.
139
Geng Y, Pan X, Xu WZ, Li B, Chen J. Plasticity and ontogenetic drift of biomass allocation in
response to above- and belowground resource availabilities in perennial herbs: a case study of
Alternanthera philoxeroides. Ecol Res. 2007;22:25560.
Ghalambor CK, McKay JK, Carroll SP, Reznick DN. Adaptive versus non-adaptive phenotypic
plasticity and the potential for contemporary adaptation in new environments. Funct Ecol.
2007;21:394407.
Hunt R. Basic growth analysis. London: Unwin Hyman Press; 1990.
McConnaughay KDM, Coleman JS. Biomass allocation in plants: ontogeny or optimality? A test
along three resource gradients. Ecology. 1999;80:258193.
Mooney HA, Kuppers M, Koch G, Gorham J, Chu C, Winner WE. Compensating effects to growth
of carbon partitioning changes in response to SO2-induced photosynthetic reduction in radish.
Oecologia. 1988;75:5026.
Newman RA. Adaptive plasticity in amphibian metamorphosis. BioScience. 1992;42:6718.
Niinemets U. Adaptive adjustments to light in foliage and whole-plant characteristics depend on
relative age in the perennial herb Leontodon hispidus. New Phytol. 2004;162:68396.
Pigliucci M. Phenotypic plasticity: beyond nature and nature. Baltimore: The John Hopkins
University Press; 2001.
Poorter H, Claudius ADM, van de Vijver CADM, Boot RGA. Growth and carbon economy of a
fast-growing and a slow-growing grass species as a dependent on nitrate supply. Plant Soil.
1994;171:21727.
Rice SA, Bazzaz FA. Quantification of plasticity of plant traits in response to light intensity:
comparing phenotypes at a common weight. Oecologia. 1989;78:5027.
Scheiner SM. Towards a more synthetic view of evolution. Am J Bot. 1999;86:1458.
Schlichting CD. The evolution of phenotypic plasticity in plants. Annu Rev Ecol Evol Syst.
1986;17:66793.
Siegal ML, Bergman A. Waddingtons canalization revisited: developmental stability and evolution. Proc Natl Acad Sci U S A. 2002;99:1052832.
Silvertown J, Charlesworth D. Introduction to plant population biology. 4th ed. Oxford: Blackwell
Science; 2001.
Sultan SE. Evolutionary implications of phenotypic plasticity in plants. Evol Biol.
1987;21:12778.
Varshney CK, Garg JK, Lauenroth WK, Heitschmidt RK. Plant responses to sulfur dioxide
pollution. Crit Rev Environ Control. 1979;9:2750.
Via S. Adaptive phenotypic plasticity: target or by-product of selection in a variable environment?
Am Nat. 1993;142:35265.
Vogel S. Sun leaves and shade leaves: differences in convective heat dissipation. Ecology.
1968;49:12034.
Waddington CH. Canalization of development and the inheritance of acquired characters. Nature.
1942;15:5635.
Walbot V. Sources and consequences of phenotypic and genotypic plasticity in flowering plants.
Trends Plant Sci. 1996;1:2732.
Weiner J. Allocation, plasticity and allometry in plants. Perspect Plant Ecol Evol Syst.
2004;6:20715.
West-Eberhard MJ. Phenotypic plasticity and the origins of diversity. Annu Rev Ecol Syst.
1989;20:24978.
West-Eberhard MJ. Developmental plasticity and evolution. New York: Oxford University Press;
2003.
White J. The plant as a metapopulation. Annu Rev Ecol Syst. 1979;10:10945.
Whitman DW, Agrawal AA. What is phenotypic plasticity and why is it important? In: Whitman
DW, Ananthakrishnan TN, editors. Phenotypic plasticity of insects: mechanisms and consequences. Enfield: Science Publishers; 2009. p. 163.
Wright SD, McConnaughay KDM. Interpreting phenotypic plasticity: the importance of ontogeny.
Plant Species Biol. 2002;17:11931.
140
Further Reading
Bernacchi CJ, Coleman JS, Bazzaz FA, McConnaughay KDM. Biomass allocation in old-field
annual species grown in elevated CO2 environments: no evidence for optimal partitioning.
Glob Change Biol. 2000;8:85563.
Bernacchi CJ, Thompson JN, Coleman JS, McConnaughay KDM. Allometric analysis reveals
relatively little variation in nitrogen versus biomass accrual in four plant species exposed to
varying light, nutrients, water and CO2. Plant Cell Environ. 2007;30:121622.
Caldwell MM, Pearcy RW. Exploitation of environmental heterogeneity by plants: ecophysiological processes above- and belowground. London: Academic; 1994.
Chevin L, Lande R, Mace GM. Adaptation, plasticity, and extinction in a changing environment:
towards a predictive theory. PLoS Biol. 2010;8:e1000357. doi:10.1371/journal.pbio.1000357.
Coen E. The art of genes: how organisms make themselves. New York: Oxford University Press;
1999.
Diggle PK. A developmental morphologists perspective on plasticity. Evol Ecol.
2002;16:26783.
Gianoli E, Valladares F. Studying phenotypic plasticity: the advantage of a broach approach. Biol
J Linn Soc. 2012;105:17.
Hallgrmsson B, Hall BK. Variation: a central concept in biology. San Diego: Academic; 2005.
Hodge A. The plastic plant: root responses to heterogeneous supplies of nutrients. New Phytol.
2004;162:924.
Huber H, Lukcs S, Watson MA. Spatial structure of stoloniferous herbs: an interplay between
structural and blue-print, ontogeny and phenotypic plasticity. Plant Ecol. 1999;141:10715.
Leyser O, Day S. Mechanisms of plant development. Oxford: Blackwell; 2003.
Matesanz S, Gianoli E, Valladares F. Global change and the evolution of phenotypic plasticity in
plants. Ann NY Acad Sci. 2010;1206:3555.
McCarthy MC, Enquist BJ. Consistency between an allometric approach and optimal partitioning
theory in global patterns of plant biomass allocation. Funct Ecol. 2007;21:71320.
McConnaughay KDM, Coleman JS. Can plants track changes in nutrient availability via changes
in biomass partitioning? Plant and Soil. 2008;202:2019.
Miner BG, Sultan SE, Morgan SG, Padilla DK, Relyea RA. Ecological consequences of phenotypic plasticity. Trends Ecol Evol. 2005;20:68592.
Mooney HA, Winner WE, Pell EJ. Response of plant to multiple stresses. London: Academic;
1991.
Moriuchi KS, Winn AA. Relationships among growth, development and plastic response to
environmental quality in a perennial plant. New Phytol. 2005;166:14958.
Muller I, Schmid B, Weiner J. The effect of nutrient availability on biomass allocation patterns in
27 species of herbaceous plants. Perspect Plant Ecol Evol Syst. 2000;3:11527.
Novoplansky A. Developmental plasticity in plants: implications of noncognitive behavior. Evol
Ecol. 2002;16:17788.
Pigliucci M. Evolution of phenotypic plasticity: where are we now? Trends Ecol Evol.
2005;20:4815.
Pigliucci M, Hayden K. Phenotypic plasticity is the major determinant of changes in phenotypic
integration in Arabidopsis. New Phytol. 2001;152:41930.
Pigliucci M, Murren CJ, Schlichting CD. Phenotypic plasticity and evolution by genetic assimilation. J Exp Biol. 2006;209:23627.
141
Porter JR. A modular approach to plant growth analysis. I. Theory and principles. New Phytol.
1983;94:18390.
Porter JR, Lawlor DW. Plant growth interactions with nutrition and environment. Cambridge:
Cambridge University Press; 1991.
Reekie EG, Bazzaz FA. Reproductive allocation in plants. London: Academic; 2005.
Smith JM, Burian R, Kaufman S, Alberch P, Campbell J, Goodwin B, Lande R, Raup D, Wolpert
L. Developmental constraints and evolution. Q Rev Biol. 1985;60:26587.
Stanton ML, Roy BA, Thiede DA. Evolution in stressful environments I: phenotypic variability,
phenotypic selection, and response to selection in five distinct environmental stress. Evolution.
2000;54:93111.
Steinger T, Roy BA, Stanton ML. Evolution in stressful environments II: adaptive value and costs
of plasticity in response to low light in Sinapis arvensis. J Evol Biol. 2003;16:31323.
Sultan SE. Phenotypic plasticity for plant development, function and life history. Trends Plant Sci.
2000;5:53742.
Sultan SE, Bazzaz FA. Phenotypic plasticity in Polygonum persicaria. III. The evolution of
ecological breadth for nutrient environment. Evolution. 1993;47:105071.
Valladares F, Sanchez-Gomez D, Zavala MA. Quantitative estimation of phenotypic plasticity:
bridging the gap between evolutionary concept and its ecological applications. J Ecol.
2006;94:110316.
Valladares F, Gianoli E, Gomez JM. Ecological limits to plant phenotypic plasticity. New Phytol.
2007;176:74963.
Via S, Lande R. Genotype-environment interaction and the evolution of phenotypic plasticity.
Evolution. 1985;39:50522.
Via S, Gomulkiewicz R, De Jong G, Scheiner SM, Schlichting CD, Van Tienderen PH. Adaptive
phenotypic plasticity: consensus and controversy. Trends Ecol Evol. 1995;10:2127.
Contents
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
A Primer on Plant-Herbivore Evolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Techniques and Analyses for Testing Classic Macroevolutionary Hypotheses . . . . . . . . . . .
A Framework for Explaining the Diversity and Function of Secondary Metabolites . . . .
Ecological Costs, Trade-Offs, and the Emergence of Plant Defense Theories . . . . . . . . . . . .
This Aint a Scene, Its an Arms Race . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Fitting Plants with Weapons in the Form of Chemicals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Advantages of Chemical Mixtures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Spatiotemporal Patterns of Plant Secondary Chemistry Alter Herbivore Performance . . .
Running to Stay in the Same Place . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Basis of Plant Selection and Evolution of Feeding Deterrents . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Insect Detoxification Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Sequestration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Induced Responses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Regulation of Costly Defenses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Plant Perception and Signal Transduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Direct Inducible Defenses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Indirect Inducible Defenses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Future Directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
144
146
146
147
148
152
153
157
158
161
162
163
164
165
165
167
169
170
171
172
Abstract
143
144
A.M. Trowbridge
It takes all the running you can do to keep in the same place.
The Red Queen to Alice
Through the Looking Glass (1960)
Introduction
Plants and their associated insect herbivores account for more than half of all
described species and interactions between these organisms are among the most
dominant relationships in nature. Terrestrial plants serve as the primary food source
for more than one million insect species scattered across diverse taxa, and these insect
herbivores ingest >20 % of annual net primary productivity (Schoonhoven
et al. 2005). Insects have developed diverse feeding strategies to obtain nutrients
from their host plants, yet plants have not remained passive in the face of these
attacks. Rather, plants have developed constitutive and dynamic forms of both
physical and chemical resistance over evolutionary time to mitigate herbivory.
These plant traits consequently influence the evolutionary trajectories of herbivores,
thus resulting in the reciprocal evolution of herbivore countermeasures to thwart
defenses. Ehrlich and Raven (1964) famously coined this phenomenon as coevolution, a concept that serves as a framework for the discussion within this chapter.
Beginning in the Early Devonian, and followed by a more extensive pulse at the
Mississippian-Pennsylvanian boundary, plants have been exposed to herbivory,
resulting in a vast array of trophic connections and evolutionary radiations. The
chemical interactions that exist between plants and herbivores have long been
documented, but the field of chemical ecology has experienced substantial developments over the past 40 years for a number of reasons including (1) increasingly
successful identifications of organic molecules, (2) a merging of state-of-the-art
chemical techniques with a desire to understand complex biological systems, and
(3) the awareness that secondary metabolites play a significant role in
complex multitrophic interactions (Harborne 1997). These advancements, biological
phenomena to the development and merging of technologies for studying chemically
mediated biological phenomena, have led to exciting opportunities in the field,
creating an area ripe for integrative research and important ecological discoveries.
It was the seminal work of Fraenkel (1959) that highlighted the fact that
secondary metabolites (compounds not directly involved in growth or reproduction)
were not simply to be considered waste products of a plants primary metabolism
145
but that there may be ecological and evolutionary reasons for the existence of the
overwhelming chemical diversity of these compounds, namely, defense against
pathogens and herbivores. The composition of secondary metabolites in plants
varies not only across different plant taxa but can fluctuate substantially among
different populations, between individuals, among different organs, across developmental stages, and under varying environmental conditions. The dynamic nature
of secondary metabolism has led to the development of a number of theories
describing the various factors and selection pressures responsible for the qualitative
and quantitative patterns of defense compounds observed today. Thus, it remains
essential to explain the patterns of diversity and the distribution of various classes
of secondary metabolites to mechanistically understand how they function in plant
defense against herbivores.
Much research has demonstrated the role of plant secondary metabolites as
defense mechanisms against insect feeding via direct toxicity, reducing digestibility,
deterring feeding, and attracting the natural enemies of the herbivore. Similar to
functional analyses of other plant traits, understanding the evolution of chemical
defenses is evaluated in the context of plant fitness; thus, the term defense is usually
reserved for traits that increase plant fitness while resistance refers to a trait that
reduces herbivore preference, survival, and/or fecundity (Karban and Baldwin 1997).
The distinction between these terms is important, and while few studies have truly
demonstrated the defensive role of secondary compounds via observed increases in
plant fitness, this chapter uses this term interchangeably with resistance, assuming the
traits they describe result in a reduction in the impact of herbivores. While the
potential selective pressures responsible for shaping defensive chemistry are
discussed, the use of these terms (i.e., defense and resistance) is not meant to make
any assumptions as to the driving forces behind the evolution of plant chemical traits.
Suites of plant secondary compounds that confer resistance to insects can be
controlled genetically and expressed constitutively, providing a relatively constant
barrier to attack, or they can be induced, produced in response to tissue damage.
Plants possess the ability to assess stimuli in their environment and respond accordingly via several different potential modes of signaling (e.g., chemical, electrical,
etc.), resulting in both general and herbivore-specific chemical responses. Furthermore, plants can respond to herbivory not only by altering the concentrations of
defense compounds within their tissues but also by actively releasing volatile organic
compounds (VOCs) into the environment. These volatiles serve as an indirect defense
by recruiting the natural enemies of a plants associated herbivore(s). As such, recent
research is moving beyond chemically mediated interactions between pairs of species
and beginning to focus on large-scale multitrophic effects of plant secondary metabolites at multiple scales in time and space in an effort to understand their significance
at the community and ecosystem level.
Although many plant secondary compounds have been shown to have antagonistic effects on herbivores, insects have also evolved counteradaptations to avoid,
tolerate, and detoxify defense chemicals as well as use them to their benefit in
response to predation, disease, and environmental variability. Many studies have
stressed the importance of plant chemistry in driving the evolution of herbivore
146
A.M. Trowbridge
preference and thus herbivore host ranges. Furthermore, the distributions of secondary compounds among plant taxa have been used as evidence of their defensive
roles and their involvement in what has been termed a biochemical coevolutionary
arms race. However, plant secondary metabolites not only respond to insect
feeding but are also influenced by environmental variability. Changes in secondary
chemistry associated with fluctuating environmental conditions and observed
global changes to atmospheric composition can lead to altered trophic interactions
and may have important feedbacks on ecosystem function, a topic beyond the scope
of this chapter. To summarize, plant chemical responses are controlled by complex
and interacting abiotic and biotic factors with cascading effects on food webs,
communities, ecosystems, and atmospheric composition. Thus, a comprehensive
understanding of the role of secondary compounds on ecosystems, including underinvestigated multitrophic interactions, requires a multidisciplinary experimental
approach that combines inference from evolution, chemistry, organismal science,
and ecology.
147
molecular evolution (often termed the molecular clock) but are also calibrated to
fossil records. In other words, not all insights on past plant-insect associations can
be gleaned from phylogenetic analyses of modern taxa alone. The fossil record
provides invaluable insight into the role of plants and insects in past ecosystems by
offering data on the intensity of herbivore attack, recognizable damage types, levels
of host specialization, and temporal trends of herbivore pressure on plant lineages.
As will be discussed in more detail below, combining phylogenetic history, fossil
record patterns, and manipulative field experiments will lead us into a new era of
understanding the evolution and ecological significance of plant defenses.
148
A.M. Trowbridge
explored later in this chapter. Despite these strong theoretical foundations and years
of empirical work, understanding the ecological and evolutionary processes that
lead to variation in plant defenses among species remains a significant challenge.
To demonstrate reciprocal coevolutionary adaptations requires information on
function and selection; however, these data can be difficult to obtain and few studies
have rigorously tested this hypothesis.
Beyond explaining the diversity of plant chemical compounds, the role these
compounds play in mediating plant-herbivore interactions and the development of
host specificity among phytophagous insects remains a highly debated topic in
chemical ecology. Since Fraenkels (1959) article, the past 50 years have seen the
development, testing, and modeling of the ecological roles of secondary compounds, and the major functional roles of these compounds have been broadly
acknowledged. Secondary metabolites are heritable and produced as a result of the
selective advantages that they confer, which is consistent with their sophisticated
structures, complex mechanisms of action, and multiple observable functions in
nature. While the primary function of secondary compounds appears to be that of
defense, in many instances defensive activities coincide with the same compounds
also serving as herbivore attractants as well as transport, storage, and/or signaling
molecules. However, the possibility that the evolution of function can occur
through insect adaptation allows for much faster and more complicated evolutionary changes, a topic that will be discussed in more detail later in this chapter. In
light of the diversity of secondary compounds, their myriad ecological and physiological roles, and the speed with which those roles change, a better understanding
of the regulation of secondary metabolite production (i.e., gene expression) will
offer welcomed insight into controls over the quantities of secondary compounds
and changes in their function through time.
149
Thus, natural selection is expected to favor plants that posses a composition and
concentration of defense compounds that not only maximize diversity but also
minimize costs (Jones and Firn 1991).
The idea that there must be a cost associated with the production of secondary
compounds, or defense, was first put within an optimality framework by McKey
(1974), who suggested that an increase in reproduction resulting from an increased
allocation of resources to defense is likely due to a plant maintaining a certain level
of biomass (e.g., forgoing attack by pests or pathogens). Applying this concept to
risk assessment theory, Feeny (1976) developed the apparency model, where plants
that are noticeable to herbivores are more likely to invest in defense in contrast to
unapparent or ephemeral plants. However, assessing the apparency of a plant can be
subjective; rather than focusing on how apparent plants are to particular herbivores,
Janzen (1974) suggested that slow-growing plants, particularly those inhabiting
resource-poor environments, should invest heavily in chemical defenses due to the
value of each leaf to the plant. Plants growing in resource-rich soils would thus be
less likely to invest in defense as they would be able to grow faster and better
tolerate herbivory. Similarly, the Resource Availability Hypothesis set forth by
Coley et al. (1985) postulated that abiotic resources (e.g., high-resource light gaps)
are the driving factor behind plant evolutionary strategies or syndromes, with
escape or pioneer species having few chemical defenses but rapid leaf expansion
and low nutritional quality and defense species (understory) having high levels of
chemical defenses. Around the same time, the Carbon/Nutrient Balance Hypothesis
was developed (Bryant et al. 1983) and described how the supply of carbon
and nutrients in the environment influences the production of plant defenses.
Namely, if the C:N ratio acquired by a plant controls allocation of resources
to plant functions, carbon-based defenses will be produced under nitrogen-poor
conditions and more nitrogen-based defenses synthesized when carbon is limited.
In the 1990s, Herms and Mattson (1992) offered a synthesis and expansion of
the Growth-Differentiation Balance Hypothesis (Loomis 1932), stating that
plant defenses are a result of a trade-off between growth and differentiation (i.e.,
processes that enhance the structure or function of existing cells) and a plant will
only produce chemical defenses when sufficient energy is available from
photosynthesis.
While each of the above hypotheses have been cited at one time or another as the
theoretical basis for published studies on plant defense, considerable confusion
remains, namely, due to the fact that (1) there is a large diversity of secondary
metabolite structure and function (Table 1), (2) the hypotheses are not mutually
exclusive and are difficult to test, and (3) contradictory results have led to the
perception that there is no tangible theory of plant defense. For a more detailed
account of plant defense hypotheses, see Stamp (2003). While each hypothesis and
model study system has contributed to our current understanding of plant defenses,
they also demonstrate how unrelated plant species have converged evolutionarily
on suites of similar defense strategies.
Much evidence suggests that sets of traits have evolved independently to maximize fitness under given environmental and ecological conditions. For example,
Solanaceae
Papaveraceae
Apocynaceae
Ranunculaceae
Ubiquitous
Proteinase inhibitors
Fabaceae
Rosaceae
Linaceae
Compositae
Brassicaceae
Tropaeloaceae
Cappridaceae
Common plant
families
Fabaceae
Poaceae
Atkaloids
Glucosinolates
Cyanogenic
glycosides
Chemical class
Non-protein amino
acids
Sinigrina
Amygdalina
Linamarin
Dhurrin
Example(s)
L-canavaninea
L-DOPA
Caffeine
Atropine
Nicotinea
Strychnine
Coniine
Prevent degradation and turnover of anti-nutritional or toxic Serine
proteins
Cysteinea
Aspartic
Toxicity of HCN
Feeding deterrent
Modes of defense
Mimick protein amino acids
Interfere with enzymes and neurotransmitters
150
A.M. Trowbridge
Ubiquitous
Phenotics
Ubiquitous
Terpenoids
-pinenea
(E)--farnesene
Avenacoside-B
Digitoxin
Squalene
Retinol
Capsaicin
Salicylic acid
Quercetin
Gallic acida
Flavone
Psoralen
Angelicin
6
Evolutionary Ecology of Chemically Mediated Plant-Insect Interactions
151
152
A.M. Trowbridge
different plant lineages have evolved the ability to make the same specialized
metabolites present in other lineages or make different compounds that fulfill the
same functional role. There are a number of reasons that multiple resistance traits
may evolve together and repeatedly across species. First, most plants are subject to
multiple attackers, with specific traits negatively impacting particular pests. Thus,
diversifying resistance strategies will likely increase defense against a large number
of potential herbivores. In addition, multiple resistance traits may be adaptive
considering that some traits, while conferring defense under some circumstances,
may fail to provide resistance under another set of ecological conditions. Finally,
defensive synergism may provide higher levels of resistance than any single
defense strategy alone, although evidence of this phenomenon remains scant.
That herbivory imposes natural selection on plants, particularly in terms of the
defensive function of plant secondary metabolites, is well documented. However,
not all secondary compounds are necessarily used for plant defense, and phylogenetic comparisons can elucidate the convergent evolution of suites of plant features
or defense syndromes in response to particular herbivores and/or the environment
(Agrawal and Fishbein 2006). The observed parallelism begs the question as to
whether variation among plant taxa is mostly the result of shared biosynthetic
pathways and minor genetic changes (i.e., similar phenotypic origins) or the result
of differential histories and selection pressures. Unfortunately, the study of convergent evolution in plant defense chemistry is limited by (1) an incomplete
knowledge of the secondary metabolites within each plant species (where the
200,000 identified to date is likely a gross underestimate) and (2) a lack of
knowledge regarding the genes and biosynthetic pathways responsible for the
production of these compounds. Studying the modes of action and ecological
roles of different classes of secondary compounds will offer insights into the
evolution of plant-insect interactions.
153
The role of secondary compounds has been extensively explored within a coevolutionary framework, particularly the idea of plants possessing a chemical armory
to avoid being overeaten and selective pressures in the form of insect feeding
requirements. To complement the ideas and theories presented in the previous
section, this chapter continues to discuss plant-insect interactions as mediated by
secondary compounds in the context of biochemical coevolutionary theory, but
with a more specialized focus on the types of secondary compounds produced, their
associated costs, the specificity of their production, and the impacts they have on
herbivores. The following section offers evidence and insight into the theories
presented on the continuing coevolutionary arms race between plants and insects
for mutual survival and the patterns of host utilization and diversity of plant
secondary chemistry observed today.
154
A.M. Trowbridge
Other nonprotein amino acids, such as L-DOPA, have been shown to harm insects
by interfering with essential enzymes, such as those responsible for the hardening
and darkening of the insect cuticle. Yet other nonprotein amino acids mimic
neurotransmitters (e.g., dopamine, norepinephrine), resulting in abnormal growth
and development. While these compounds are fairly effective in defending
plants against herbivores, there is a risk in deploying these nitrogen-rich
compounds as defense agents. Some species have developed the ability to detoxify
these compounds, converting them to usable forms of nitrogen, which can be an
extremely limiting nutrient in many terrestrial ecosystems and particularly for
insects.
Cyanogenic Glycosides
While cyanogenic glycoside are not themselves toxic, when enzymatically broken
down, they release hydrogen cyanide (HCN), which affects the terminal cytochrome oxidase system in the mitochondrial respiratory pathway, resulting in
oxygen starvation and death. To prevent autotoxicity, the plant must take precautions during biosynthesis, forming multienzyme complexes which prevent the
release of harmful intermediates. Following their production, plants then store these
N-containing substances as inactive glycosides (a molecule in which a sugar is
bound to a noncarbohydrate structure) in the vacuole separate from the cytoplasmic
hydrolases (-glucosidases and -hydroxynitrilelyases). Upon herbivore feeding,
the cell structures are ruptured, including the vacuole, allowing the two substances
to interact, resulting in the cleaving of the aglycone moiety and the conversion to
HCN. Approximately 60 variations of cyanogenic glycosides have been identified
and are characteristically found in more than 2,600 plant species, including ferns,
gymnosperms, and angiosperms. Cyanogenic glycosides have been extracted from
almonds and the fruits of the Rosaceae family (e.g., cherries, apples, plums,
peaches, raspberries) and in several important crops, such as cassava (Manihot
esculenta), sorghum (Sorghum bicolor), and barley (Hordeum vulgare). Despite the
effective toxicity of HCN, its effect on herbivores is dosage dependent, as are most
defense compounds, and some specialists are capable of tolerating relatively high
levels of HCN. Furthermore, recent studies have suggested that the primary defensive role of cyanogenic glycosides does not appear to be its toxicity, but rather its
ability to serve as an effective feeding deterrent due to its bitter taste.
Glucosinolates
Close to 150 different glucosinolates, or mustard oil glycosides, have been identified within the Brassicaceae, Capparidaceae, and Tropaeolaceae plant families.
Glucosinolates are biosynthetically related to cyanogenic glycosides as both are
spatially separated from their hydrolyzing enzyme, in this case a thioglucosidase
myrosinase. Similar to the production of HCN from cyanogenic glycosides, the
enzyme and glucosinolate substrate come into contact upon tissue damage from
herbivory, and the unstable aglycones are released resulting in various active
compounds including nitriles and isothiocyanates. The latter hydrolysis product
affects herbivores by reacting spontaneously with compounds containing unshared
155
pairs of electrons, mainly proteins and nucleic acids, making them inactive. However, the role of glucosinolates as defensive compounds is complicated by the
extreme variation in their composition and concentration within species, between
plant tissues, and across ontogenetic stages, providing both a challenge and
opportunity for specialist and generalist insects. For example, while significant
negative correlations have been shown between glucosinolate content and insect
fecundity, several specialist species preferentially feed on Brassicaceae,
using antennal receptors to locate their preferred hosts by the presence of
glucosinolates.
Alkaloids
Alkaloids are one of the most structurally diverse groups of N-containing secondary
compounds and are present in ~20 % of higher plant families, including Solanaceae,
Papaveraceae, Apocynaceae, and Ranunculaceae. More than 12,000 alkaloids
have been identified to date, and these can be subdivided into more than 20 different
classes including pyrrolidines, tropanes, piperidines, and pyridines. With individual
alkaloids having the ability to carry out multiple functions, it is not surprising
that these compounds can exhibit a variety of deleterious effects on metabolic
function and physiology by affecting enzymes, inhibiting DNA synthesis and
repair, and affecting the nervous system. In addition to the famous use of alkaloids
extracted from hemlock to put the philosopher Socrates to death, other typical
alkaloids include caffeine, atropine, and nicotine. Besides their well-known effects
on vertebrates, including humans, alkaloids act as natural defense compounds
by paralyzing and having toxic effects on herbivores, such as targeting insect
postsynaptic receptors, as in the case of nicotine.
Proteinase Inhibitors
Proteinase inhibitors do just that, inhibit different types of proteinases that occur in
the herbivore gut, ultimately serving as anti-digestive proteins. Proteinase inhibitors interact with the active site of target proteases, attenuating protein processing
and turnover by causing enzymes to become inactive, thus preventing the degradation of anti-nutritional or toxic proteins and interfering with digestion in the gut
to prevent effective nutrient utilization. Plants contain a variety of proteinase
inhibitors (e.g., serine, cysteine, aspartic), and the various classes are identified
by the structure of their polypeptide backbone. Some proteinase inhibitors are
found constitutively in seeds and tubers, likely because the integrity of these organs
is essential for survival. Herbivore attack can also induce proteinase inhibitor gene
expression, both locally and systemically. While the effects of proteinase inhibitors
on herbivore mortality or performance are relatively minor, even small effects on
development or fecundity may be ecologically relevant.
Terpenoids
Terpenoids are ubiquitous across plant families, with over 22,000 of these lipophilic compounds having been described, and play multiple roles in plant defense.
Terpenoids share a common biosynthetic origin and are synthesized from
156
A.M. Trowbridge
Phenolics
Similar to terpenoids, phenolics are a large, ubiquitous group of carbon-based
secondary compounds, of which over 9,000 have identified, and include a wide
variety of subclasses such as flavonoids, tannins, lignin, and furanocoumarins.
Phenolics consist of a hydroxyl group ( OH) bonded directly to an aromatic
hydrocarbon group and are classified based on the number of phenol units in the
molecule. Plants store phenolics in the vacuole. Some common and naturally
occurring phenolic compounds include cannabinoids found in Cannabis spp.,
capsaicin in chili peppers (Capsicum spp.), and salicylic acid from Salix spp.,
which is used to produce aspirin. Phenolics can have negative effects on
non-adapted insects, likely due to oxidative mechanisms in the midgut that result
in the formation of superoxide radicals and other reactive oxygen species that can
lead to protein and lipid peroxidation. Flavonoids play a variety of biological
activities (e.g., phytoalexins, detoxifying agents, UV filters, allelochemicals, etc.),
protecting plants from different biotic and abiotic stresses. However, these multiple
roles make the interpretation of experimental results regarding flavonoids
rather difficult when trying to elucidate their primary role in plant resistance.
Tannins have also been shown to be toxic to insects due to their ability to bind to
salivary proteins and digestive enzymes (e.g., trypsin and chymotrypsin), resulting
in protein inactivation, the inability to gain weight, and death. Another group of
phenolic compounds, furanocoumarins, is found primarily in species of the
157
158
A.M. Trowbridge
159
structures as glandular hairs or trichomes, leaf waxes, leaf resins, latex, and in the
vacuoles of epidermal cells. While the Optimal Defense Theory predicts roots to
have lower levels of defense compounds compared to shoots due to a lower
probability of attack, recent work has shown roots to contain a wide range of
secondary compounds at relatively high concentrations. In fact, root herbivores
have been shown to do as much or even more damage to wild plants as aboveground
feeding insects. These findings have, among others, resulted in a recent body of
literature focused on how plant secondary compounds mediate interactions between
root and shoot herbivores and vice versa. While aboveground herbivory can alter
belowground interactions via changes in root chemistry, leaf damage has also been
shown to induce the production of secondary compounds in petals, nectar, and
pollen, which may defend the plant against florivores but also deter pollinator
visitation and potentially plant fitness (see section Induced Responses below).
Understanding these differential pressures and allocation strategies will ultimately
aid in our understanding of how organisms adapt, evolve, and express particular
traits.
In addition to variation within plants, there is also considerable variability in
leaf secondary chemistry among different individuals within a population. While
trees that exhibit higher levels of defense compounds tend to experience lower
levels of herbivore damage, this can come with a significant cost to the tree in
terms of biomass (fewer leaves) resulting in cascading negative effects on nutrient acquisition and fitness. In some cases, patchiness can benefit the herbivore if it
results in particular plant species expressing low levels of defense compounds
clumped together in space (Moore and DeGabriel 2012). However, in other
instances, an herbivores search for a suitable host can be similar to trying to
find a needle in a haystack, with the herbivore being forced to spend more
time moving within the canopy and thus increasing its chance of being
predated upon by its natural enemies. Furthermore, the heterogeneous chemical
environment of the canopy can result in patches of plants defended by different
secondary compounds, which can decrease the ability of herbivores that rely on
mixed diets to ameliorate the detrimental effects of some secondary compounds.
Canopy and landscape heterogeneity in levels of defense compounds can also
lead to associational resistance, where plant susceptibility to insect pests is
influenced by the quality and proximity of neighboring plants (Barbosa
et al. 2009). Associational resistance can occur if herbivores select hosts at the
patch scale and plants gain additional resistance if neighboring plants are unpalatable. However, having unpalatable neighbors can also result in associational
susceptibility if herbivores forage at an individual plant scale, where the
contrast in nutrient value and levels of defense become more apparent. However,
whether or not variability in defense compounds among individuals results in
associational resistance or susceptibility depends on the herbivores specific
foraging movements and host location strategies, a topic that requires coupling
observations in natural landscapes with foraging theory (see Moore and
DeGabriel 2012).
160
A.M. Trowbridge
Temporal Heterogeneity
Secondary metabolites can change in response to a plants developmental trajectory (ontogeny) and to seasonal conditions (e.g., water availability, temperature,
photoperiod), with subsequent effects on a plants physiology and metabolism.
Meta-analyses performed by Barton and Koricheva (2010) revealed general
patterns for defense compounds, particularly that concentrations of secondary
metabolites tend to increase during seedling growth but decrease during leaf
development. Furthermore, they found that most metabolites remain relatively
stable in mature leaves through the season, with some changes in composition
for specific compounds (e.g., tannins and lignin). Temporal shifts in patterns of
secondary compounds may result from a number of different mechanisms,
including a potential dilution effect as other metabolites accumulate in greater
concentrations over time, translocation of compounds from one plant tissue or
organ to another, time lags associated with the differentiation of specialized
storage structures (e.g., resin ducts, trichomes), altered foliar concentrations
due to volatilization (particularly with changes in temperature), and/or catabolism (Koricheva and Barton 2012). But what are the evolutionary causes responsible for temporal changes in plant defense compounds? A number of theories
(e.g., the Growth-Differentiation Balance Hypothesis) suggest resource and
metabolic constraints, substrate-level competition, and trade-offs with other
plant processes, although little support for these mechanistic hypotheses has
emerged. Another potential explanation for the evolution of temporal changes
in plant defense compounds is that they are adaptive responses to selection
pressures imposed by herbivores. For example, a number of studies support
the Plant Apparency Hypothesis (Feeny 1976) with higher levels of secondary
compounds observed in young developing leaves and a general increase in
plant allocation to chemical defenses through ontogeny. The Optimal Defense
Theory (Rhoades 1979) is also supported with many studies showing the majority
of chemical defense present in high concentrations in young leaves, thus allocating resources to defenses proportional to the risk from herbivores and the value of
the tissue to fitness. While there is a significant amount of support for the idea that
herbivores drive temporal patterns of plant defense, other selective agents (e.g.,
pollinators, pathogens, large herbivores, plants, and the environment) may also
play a role when particular plants defense compounds are expressed.
Regardless of how temporal patterns of plant defenses evolved, it is well known
that these changes have profound effects on herbivores and the amount of damage
sustained by plants. Variation in herbivore preference and performance on leaves of
varying ages across the season may result from temporal changes in the concentration and composition of defense compounds. Likewise, changes in plant chemistry
may also affect when herbivores can feed. For instance, long-lived leaves of tropical
species are most susceptible to herbivory during the short period of leaf expansion, at
which time these leaves usually exhibit the highest concentrations of secondary
defenses (Coley et al. 1985). However, it is difficult to attribute changes in herbivore
feeding with seasonal changes in secondary compounds alone considering the
many other physiological, and thus nutritional, changes simultaneously occurring.
161
Further complicating the interpretation of how temporal changes in defense compounds alter herbivore performance is the lack of consideration for the third trophic
level. Volatile secondary metabolites indirectly serve as plant defense compounds
by attracting the natural enemies of herbivores (see section Indirect Inducible
Defenses below), and their composition and concentrations are also impacted by
leaf age and the overall developmental stage of the plant. Being more or less
attractive to parasitoids during different phases of the herbivores life cycle can
have profound effects on the effectiveness of parasitoids and result in important
consequences for current herbivore population densities and the potential for future
outbreak events.
Not only do plants exhibit an astounding level of variability in the levels of defense
compounds present in their tissues over relatively short temporal scales (e.g., a
growing season), but plant chemical defense has also been shown to respond to
temporal changes on a large environmental scale, including shifts in soil conditions
and competitive interactions among plants during succession. Herbivore communities
may also vary temporally with succession, particularly in response to altered plant
composition. Thus, it is not surprising that certain types and concentrations of plant
chemical defenses may be more prominent and effective during specific successional
stages. While plant defense theories (e.g., the Resource Availability Hypothesis)
suggest mechanisms to explain the composition and concentration of plant secondary
compounds during succession, many alternative scenarios have been observed.
For example, greater environmental heterogeneity in late successional communities
may tend to increase intraspecific variation in the quantity or quality of defense
compounds, independent of nitrogen or carbon availability relative to demand.
Unfortunately, the level of intraspecific variation in secondary chemistry among
individual plants and the mechanisms responsible for altering concentrations during
succession remains to be elucidated. A greater understanding of the specific processes
structuring these observed patterns will likely be gained by quantifying the distribution of defense compounds within natural populations.
162
A.M. Trowbridge
163
upon insect host use, it has been suggested that the development of novel insect
detoxification systems may have allowed insects to use other hosts in different plant
taxa, resulting in the radiation of local populations into new species. Thus, insects
can be polyphagous (eating any plant, such as leaf cutting ants and locusts),
oligophagous (feeding on relatively few related species belonging to one or only
a few plant genera or families, such as danaid butterflies), or monophagous (feeding
on a single plant species, such as the silkworm on mulberry leaves). Most insects
are mono- or oligophagous and have been shown to feed on obviously toxic plants,
and some even have the ability to exploit these defense compounds for their own
use (see section Sequestration below). In accordance with coevolutionary theory,
an evolving pattern of feeding deterrents has arisen across plant taxa with a trend
toward chemical complexity in structures. While the production of such a diversity
of chemical structures serves as an effective defensive function for plants, some
insects have evolved defense mechanisms to cope with the negative effects of these
compounds. In addition to the adaptive processes of insects, biochemical information has also shed light on the plasticity of plants and their energetic, metabolic, and
chemical responses to producing costly chemical defenses following herbivory (see
section Induced Responses below). Taken together, these interactions (insect
detoxification mechanisms and the relative costs and benefits of plant chemical
defenses) offer insight into the diversity of both plant and insect species and the
theory of reciprocal evolutionary interactions mediated by plant chemistry.
164
A.M. Trowbridge
Sequestration
In addition to having the ability to biochemically digest/assimilate plant defense
compounds present in their hosts, herbivores can also make use of secondary
compounds obtained from plants by storing them in their own body tissues and
integument. An impressive variety of plant defense compounds can be sequestered
by insects including alkaloids, cyanogenic glycosides, glucosinolates, and
isoprenoids. More than 250 herbivorous insect species in six orders have demonstrated the ability to sequester these metabolites from at least 40 plant families
(Opitz and M
uller 2009). The amounts of sequestered compounds found in insects
can vary dramatically, in part due to the variability of secondary metabolites present
in host plants (Nishida 2002). The amount of sequestered compounds can also differ
between insect sexes depending on their use for reproductive purposes (e.g., serving
as precursors for pheromone production, nuptial gifts or spermatophores, or offspring protection). In many Lepidoptera and Hymenoptera, sequestered compounds
can also be transferred from the leaf-chewing larval stage to the adult stage, a
strategy that can be used to the insects advantage by making adults unpalatable to
natural enemies. One of the most famous examples of defense chemicals benefiting
insects through sequestration is that of the specialist monarch butterfly
(Danaus plexippus Nymphalidae), its milkweed (Asclepias syriaca) host plant,
and its natural predator, the blue jay (Cyanocitta cristata). The milkweed
provides monarchs with cardiac glycosides, which they sequester, rendering them
poisonous to most vertebrates. Blue jays that attempt to eat monarch butterflies tend
to have a strong visceral reaction to the toxic compounds and learn to associate the
markings of the butterfly with this response, thereby learning to avoid them in the
future. However, it is important to note that the benefits gained by monarchs from
cardiac glycosides do come with a cost: monarchs are negatively affected when
feeding on milkweed plants with low nitrogen levels (having to consume more
plant tissue per day and making them more vulnerable to predation) and can also
be negatively affected by high levels of cardiac glycosides and/or latex. Thus,
the heterogeneous nature of individual plant quality within a population plays an
important role in host choice via the ability to cope with a plants defense
mechanisms.
In order to successfully sequester and exploit plant defense compounds without
inflicting harm on themselves, insects have evolved a number of interesting physical and biochemical strategies. To avoid autotoxicity, insects are forced to
165
Induced Responses
An element of plant-insect interactions that eludes a simple evolutionary description is that of induced responses. Despite the amount of attention induced resistance
has received over the past few decades, there remains a significant gap in our
knowledge pertaining to the evolution of herbivore-induced specificity of defense
strategies and perception. In addition, the heritability of herbivore-induced
responses is yet to be determined, particularly the genetic basis for hormonal
signaling, the interactions between pathways, and the selective forces that act on
these traits. In terms of the latter, the influence of induced responses across trophic
levels within communities is complex, eliciting various responses in different
arthropod species resulting in varying degrees of selection. While more work is
needed, it is clear that a community perspective is critical to understanding costs of
chemical defense syndromes and the evolution of specificity in plant induced
responses (Agrawal and Fishbein 2006)
166
A.M. Trowbridge
Fig. 1 Simplified scheme of the signal cascade involved in plant perception of herbivory and the
synthesis of secondary defense compounds. Solid lines represent systemic upregulation of genes
responsible for secondary metabolite production (leafs, flowers, and roots) and the dashed line
indicates a local response at the area of wounding (leaf level). Resulting direct and indirect
herbivore-induced defenses mediated by changes in plant chemistry are indicated across trophic
levels with colored circles. Orange circles represent direct defenses in the form of reducing
nutritive value, toxicity, and/or volatile feeding and oviposition deterrents for conspecifics, and
blue circles represent indirect defenses or the volatile attraction of the natural enemies of
herbivores, while purple circles indicate variability in the attractive or deterrent properties of
altered chemistry in distal plant organs
of existing secondary metabolites (e.g., Kessler and Baldwin 2002). Herbivoreinduced changes in plant chemistry can have either direct effects on the susceptibility of host plants to insects or can serve as attractants to the natural enemies of the
herbivore, serving as an indirect defense (see section Indirect Inducible Defenses
below). As with constitutive defenses, inducible defenses can only truly be selected
as defense systems if there is heritable variation and if the plants experience a
higher fitness by exhibiting the induced chemical response than not. However,
because a true measure of fitness is difficult to obtain, proxies must be used and
the results can be inconsistent and variable. As noted in the beginning of the
chapter, the use of the term defense makes no assumptions about selection by
herbivores, only that the trait defends the plant. Whether it evolved specifically to
do so is another issue and it is difficult to determine the specific selective factors
that shape a trait. Fortunately, phylogenetic reconstructions have begun to offer
167
insight into the validity of the coevolutionary theory and the history of induced
responses. By placing theories describing the variability of induced defenses within
the framework of costs and constraints, one can hope to begin to understand the
evolutionary development of induced chemical traits.
Investing in compounds that defend plants can be costly and plants must
allocate their finite resources among defense, growth, and reproduction as
needed. Thus, investing in the synthesis of defense compounds unnecessarily
can be directly costly in terms of fitness. However, if the compounds conferring
defense also provide other benefits, such as dissipating heat, providing structure,
etc., then the ecological cost may be relatively little in terms of plant fitness, even
in the absence of herbivores. Plants may also pay a cost in terms of preventing
autotoxicity, by synthesizing structures to safely store toxic compounds if the
defense is constitutively maintained. Plants may also incur resistance costs,
which can result from higher-level ecological interactions, such as in the case
of compounds serving as deterrents against generalist herbivores as well as
pollinators (Kessler and Baldwin 2002). The inconsistent phenotype expressed
by plants with induced defenses may benefit the plant by resulting in a lag in
insect counteradaptations, as herbivores are less likely to adapt to defenses that
are intermittently expressed as compared to those they encounter on a more
regular basis. Overall, in the absence of herbivores, it appears to be in the best
interest of the plant to avoid the aforementioned metabolic and ecological fitness
costs, favoring the evolution of inducible defenses. Despite the vast number of
induced secondary compounds described and their potential roles in plant protection, definitive proof of the particular defense function of each compound
remains to be determined in most cases.
168
A.M. Trowbridge
is the inceptins, which are produced when a plant ATP-synthase subunit is cleaved
in the midgut of the insect. Small molecular elicitors known as caeliferins can also
induce chemical responses as well as lytic enzymes, such as -glucosidase (isolated,
e.g., from Pieris brassicae), glucose oxidases (Helicoverpa zea), alkaline phosphatase (piercing whitefly Bemisia tabaci), and watery digestive enzymes from aphid
saliva. Furthermore, the feeding behaviors of insect larvae (e.g., speed, mode,
frequency) can also be differentially recognized by the plant and play an important
role in the specificity of the induced response. In addition to herbivore feeding,
plants can also perceive insects oviposition activities and express induced direct or
indirect defenses in response. Bruchins (isolated, e.g., from the oviposition fluid of
pea weevils) have been shown to result in neoplasma growth while benzyl
cyanide (isolated from the large cabbage white butterfly Pieris brassicae) can
induce the arrest of the parasitoid Trichogramma brassicae on Brussels sprout
(Brassica oleracea var. gemmifera). While relatively little is known about how
plants perceive herbivores, many small molecules have been identified in the
complex signaling networks responsible for deploying the appropriate downstream
defenses.
The use of model plants, artificially generated mutants, sequencing technologies, microarrays, and transcriptional profiling tools has greatly enhanced our
understanding of the genetic basis of plant signaling and stress response.
A number of regulatory networks have been suggested to mediate herbivoreinduced responses in plants including Ca2+ ion fluxes, mitogen-activated protein
kinases (MAPKs), jasmonic acid (JA), salicylic acid (SA), ethylene (ET), and
reactive oxygen species (ROS). Molecular and genomic tools are being used to
uncover the complexity of the induced defense signaling networks that have
evolved during the arms race between plants and their attackers. While descriptions of each of these signaling pathways is beyond the scope of this chapter,
many studies have shown JA, SA, and ET to be key players in the regulation of
the signaling cascades responsible for induced defenses. Following insect attack,
plants produce varying amounts of SA, JA, and ET, which contribute to the
specific induced defenses that are synthesized. Plants may be attacked by a
number of different insect pests, requiring regulatory mechanisms that can
adapt with the various challenges they encounter. Thus, cross talk between
induced defense signaling pathways not only provides flexibility in a plants
response due to the antagonistic or synergistic interactions between the hormones
produced, but allows for a level of specificity while minimizing energy costs.
While cross talk between pathways generally aids the plant in determining which
defense strategy to deploy, some insects have evolved to manipulate plants for
their own benefit by suppressing or adjusting the production of induced defenses.
For example, some herbivores may activate the SA-signaling pathway, which
antagonistically interacts with JA-dependent defenses, thus resulting in enhanced
insect performance. Despite the progress in identifying molecular mechanisms
responsible for interactions between defense signaling pathways, explaining the
evolution and maintenance of variation in induced responses and its effect on the
fitness of plants within complex communities remains a challenge.
169
170
A.M. Trowbridge
171
microbial colonization, and alleviating abiotic stress such as drought, UV, and heat.
As such, their role in plant evolution is dynamic. A number of adaptive explanations have been offered to address the diversity of volatiles found among and within
plant families, and it has also been suggested that natural selection exploits the
volatility of the compounds themselves and the context in which they are perceived
by herbivores and their natural enemies. Similar to foliar compounds, the precise
ecological function and evolutionary consequence of every plant volatile is not yet
known so their full contribution to plant-insect evolution has yet to be characterized. However, the importance of herbivore-induced volatiles to plant, herbivore,
and parasitoid signaling and fitness highlights their potentially important role in the
coevolution among taxonomic groups.
Future Directions
The ability of plants to chemically defend themselves against the constant
onslaught of herbivores that rely on them for food and energy has fascinated
scientists for years. Since, Fraenkel (1959), many studies have sought to describe
the variation in plant chemical defense strategies that exists among and within plant
families, primarily in the context of coevolutionary theories. Coupling phylogenetic
and molecular tools with historical biogeography, studies have shown patterns in
plant chemical defense and insect host use, including convergent evolution, and
researchers must continue to enhance the molecular and chemical toolbox and
design experiments in the context of broader ecological scales to understand larger
macroevolutionary patterns. It is also necessary to understand the trade-offs that
exist between costs of chemical defense and the benefits obtained from them to
appreciate how these strategies are selected upon and evolve. However, given the
numerous secondary compounds plants produce and the range of herbivores,
pathogens, and abiotic stresses that may select for these each chemical trait over
time, determining true defensive functions of mixtures, classes of compounds, or
even individual chemicals can be daunting. Furthermore, bioassays aimed at determining the effects of these compounds on potential pests are required and should be
coupled with other genetic methods (e.g., transcriptional profiling, mutants, genetic
knockouts, etc.) to elucidate not only the effects on herbivore performance but also
the molecular mechanisms responsible for them. Along this vein, it is critical to
identify the genetic basis for hormonal signaling and interactions between pathways
in order to link plant perception of herbivores, signaling cascades, and the production of defensive compounds with the ecological repercussions at the community
level.
Plant chemical defenses cannot be considered solely on a pairwise level with a
single herbivore but must be framed within a large community perspective considering the multitude of herbivores that plants must defend against and the myriad
higher-trophic-level interactions and environmental factors that also influence plant
traits (Fig. 1). Thus, future work should focus not only on the defensive properties
of secondary compounds in terms of affecting herbivore performance, but also on
172
A.M. Trowbridge
the responses of other insects (e.g., pollinators, parasitoids, etc.) to gain a more
comprehensive understanding of the cascading effects of plant defenses on community structure. Furthermore, the extent of the specificity of plant chemical
defenses should be taken into account, particularly induced defenses, to untangle
the primary drivers of community interactions and their role in shaping plant-insect
relationships and evolutionary trajectories. In regard to specificity, it is important to
expand some of the more conventional targeted chemical analyses (i.e., only focusing on one group of compounds) and to integrate metabolomics into plant-insect
research. Current studies may be missing other important secondary compounds that
might be contributing to a plants defense against herbivores and a more mechanistic
understanding of defense allocation in plants would be gained by linking primary
and secondary metabolic processes through metabolomics. Techniques from
metabolomics may be able to detect subtle changes in plant responses over time
offering a better idea of the temporal scales over which responses might be most
effective against insect pests. While the production of plant secondary compounds
can vary significantly over time and space, it is also influenced by a suite of abiotic
factors including changes in atmospheric CO2, O3, temperature, precipitation, nutrient availability, etc., thus having important consequences for plant defense and a
number of ecological interactions. To identify general patterns of plant defense
strategies under natural conditions, future research must focus on the interactive
effects of herbivory and climate on plant secondary production and the consequences
for insect population dynamics. Thus, the impact of climate and herbivory on more
classes of compounds must be assessed in a wider range of species (i.e., outside the
boreal and temperate zone bias) and couched within a whole ecosystem context.
Such a multifactor approach is critical to understand the impacts of predicted climate
change, insect population dynamics, and their interactions in the future.
References
Agrawal AA. Current trends in the evolutionary ecology of plant defence. Funct Ecol.
2011;25:42032.
Agrawal AA, Fishbein M. Plant defense syndromes. Ecology. 2006;87:S13249.
Barbosa P, Hines J, Kaplan I, et al. Associational resistance and associational susceptibility:
having right or wrong neighbors. Annu Rev Ecol Evol Syst. 2009;40:120.
Barton KE, Koricheva J. The ontogeny of plant defense and herbivory: characterizing general
patterns using meta-analysis. Am Nat. 2010;175:48193.
Bernays E, Chapman R. The evolution of deterrent responses in plant-feeding insects. In: Chapman RF et al., editors. Perspectives in chemoreception and behavior. New York: Springer;
1987. p. 15973.
Bernays E, Graham M. On the evolution of host specificity in phytophagous arthropods. Ecology.
1988;69:88692.
Bryant JP, Chapin III FS, Klein DR. Carbon/nutrient balance of boreal plants in relation to
vertebrate herbivory. Oikos. 1983;40:35768.
Chen M-S. Inducible direct plant defense against insect herbivores: a review. Insect Sci.
2008;15:10114.
173
Coley PD, Bryant JP, Chapin III FS. Resource availability and plant antiherbivore defense.
Science. 1985;230:8959.
Despres L, David J-P, Gallet C. The evolutionary ecology of insect resistance to plant chemicals.
Trends Ecol Evol. 2007;22:298307.
Dethier VG. Evolution of feeding preferences in phytophagous insects. Evolution. 1954;8:3354.
Dicke M, Baldwin IT. The evolutionary context for herbivore-induced plant volatiles: beyond the
cry for help. Trends in Plant Science. 2010;15:16775.
Ehrlich PR, Raven PH. Butterflies and plants: a study in coevolution. Evolution. 1964;18:586608.
Feeny P. Plant apparency and chemical defense. In: Wallace JW, Mansell RL, editors.
Biochemical interactions between plants and insects. New York: Springer; 1976. p. 140.
Fraenkel GS. The raison detre of secondary plant substances. Science. 1959;129:146670.
Futuyma DJ, Keese MC. Evolution and coevolution of plants and phytophagous arthropods. In:
Rosenthal GA, Berenbaum MR, editors. Herbivores: their interactions with secondary plant
metabolites vol II: ecological and evolutionary processes. San Diego: Academic Press; 1992.
p. 439475.
Gershenzon J, Fontana A, Burow M, et al. Mixtures of plant secondary metabolites: metabolic
origins and ecological benefits. In: Iason GR, Dicke M, Hartley SE, editors. The ecology of
plant secondary metabolites: from genes to global processes. New York: Cambridge University
Press; 2012. p. 5677.
Harborne JB. Introduction to ecological biochemistry. 4th ed. San Diego: Academic; 1997.
Haukioja E. Induction of defenses in trees. Annu Rev Entomol. 1991;36:2542.
Herms DA, Mattson WJ. The dilemma of plants: to grow or defend. Q Rev Biol. 1992;67:283335.
Howe GA, Jander G. Plant immunity to insect herbivores. Annu Rev Plant Biol. 2008;59:4166.
Janzen DH. Tropical blackwater rivers, animals, and mast fruiting by the Dipterocarpaceae.
Biotropica. 1974;6:69103.
Jones CG, Firn RD. On the evolution of plant secondary chemical diversity. Philos Trans Biol Sci.
1991;333:27380.
Karban R, Baldwin IT. Induced responses to herbivory. Chicago: Chicago University Press; 1997.
Kessler A, Baldwin IT. Plant responses to insect herbivory: the emerging molecular analysis. Annu
Rev Plant Biol. 2002;53:299328.
Koricheva J, Barton KE. Temporal changes in plant secondary metabolite production: patterns,
causes, and consequences. In: Iason GR, Dicke M, Hartley SE, editors. The ecology of plant
secondary metabolites: from genes to global processes. New York: Cambridge University
Press; 2012. p. 3455.
Loomis WE. Growth-differentiation balance vs. carbohydrate-nitrogen ratio. Proc Am Soc Hortic
Sci. 1932;29:2405.
McKey D. Adaptive patterns in alkaloid physiology. Am Nat. 1974;108:30520.
Moore B, DeGabriel JL. Integrating the effects of PSMs on vertebrate herbivores across spatial and
temporal scales. In: Iason GR, Dicke M, Hartley SE, editors. The ecology of plant secondary
metabolites: from genes to global processes. New York: Cambridge University Press; 2012.
p. 22646.
Nishida R. Sequestration of defensive substances from plants by lepidoptera. Annu Rev Entomol.
2002;47:5792.
Opitz SEW, Muller C. Plant chemistry and insect sequestration. Chemoecology. 2009;19:11754.
Rhoades DF. Evolution of plant chemical defense against herbivores. In: Rosenthal GA,
Janzen DH, editors. Herbivores: their interaction with secondary plant metabolites.
New York: Academic; 1979. p. 354.
Schoonhoven LM, van Loon JJA, Dicke M. Insect-plant biology. Oxford: Oxford University Press;
2005.
Stamp N. Out of the quagmire of plant defense hypotheses. Q Rev Biol. 2003;78:2355.
Wu J, Baldwin IT. New insights into plant responses to the attack from insect herbivores. Annu
Rev Genet. 2010;44:124.
174
A.M. Trowbridge
Further Reading
Agrawal AA. Natural selection on common milkweed (Asclepias syriaca) by a community of
specialized insect herbivores. Evolut Ecol Res. 2005;7:65167.
Agrawal AA, Lau JA, Hamback PA. Community heterogeneity and the evolution of interactions
between plants and insect herbivores. Q Rev Biol. 2006;81:34976.
Agrawal AA, Conner JK, Rasmann S. Tradeoffs and negative correlations in evolutionary ecology.
In: Bell M, Eanes W, Futuyma D, Levinton J, editors. Evolution after Darwin: the first
150 years. Sunderland: Sinauer Associates; 2010. p. 24368.
Arnason JT, Bernards M. Impact of constitutive plant natural products on herbivores and pathogens. Can J Zool. 2010;88:61527.
Ayres MP, Clausen TP, MacLean SEJ, et al. Diversity of structure and antiherbivore activity in
condensed tannins. Ecology. 1997;78:1696712.
Bailey JK, Schweitzer JA, Rehill BJ, et al. Rapid shifts in the chemical composition of aspen
forests: an introduced herbivore as an agent of natural selection. Biol Invasions.
2007;9:71522.
Berenbaum M. Toxicity of a furanocoumarin to armyworms: a case of biosynthetic escape from
insect herbivores. Science. 1978;201:5324.
Berenbaum M. Patterns of furanocoumarin distribution and insect herbivory in the Umbelliferae:
plant chemistry and community structure. Ecology. 1981;62:125466.
Berenbaum M. Coumarins and caterpillars: a case for coevolution. Evolution. 1983;37:16379.
Berenbaum MC. The expected effect of a combination of agents: the general solution. J Theor
Biol. 1985;114:41331.
Berenbaum MR, Zangerl AR. Furanocoumarin metabolism in Papilio polyxenes: biochemistry,
genetic variability, and ecological significance. Oecologia. 1993;95:3705.
Berenbaum MR, Nitao JK, Zangerl AR. Adaptive significance of furanocoumarin diversity in
Pastinaca sativa (Apiaceae). J Chem Ecol. 1991;17:20715.
Berenbaum MR, Favret C, Schuler MA. On definingkey innovations in an adaptive radiation:
cytochrome P450s and papilionidae. Am Nat. 1996;148:S13955.
Bergvall UA, Rautio P, Kesti K, et al. Associational effects of plant defences in relation to withinand between-patch food choice by a mammalian herbivore: neighbour contrast susceptibility
and defence. Oecologia. 2006;147:25360.
Bernasconi ML, Turlings TCJ, Ambrosetti L, et al. Herbivore-induced emissions of maize
volatiles repel the corn leaf aphid, shape Rhopalosiphum maidis. Entomol Exp Appl.
1998;87:13342.
Bowers MD. The evolution of unpalatability and the cost of chemical defense in insects. In:
Roitberg BD, Isman MG, editors. Insect chemical ecology: an evolutionary approach.
New York: Chapman and Hall; 1992. p. 21644.
Castaneda LE, Figueroa CC, Fuentes-Contreras E, et al. Energetic costs of detoxification systems
in herbivores feeding on chemically defended host plants: a correlational study in the grain
aphid, Sitobion avenae. J Exp Biol. 2009;212:118590.
Close DC, McArthur C. Rethinking the role of many plant phenolicsprotection from
photodamage not herbivores? Oikos. 2002;99:16672.
Coley PD. Herbivory and defensive characteristics of tree species in a lowland tropical forest. Ecol
Monogr. 1983;53:20934.
De Moraes CM, Lewis WJ, Pare PW, et al. Herbivore-infested plants selectively attract parasitoids. Lett Nat. 1998;393:5703.
De Moraes CM, Mescher MC, Tumlinson JH. Caterpillar-induced nocturnal plant volatiles repel
conspecific females. Nature. 2001;410:57780.
Degenhardt J, Kollner TG, Gershenzon J. Monoterpene and sesquiterpene synthases and the origin
of terpene skeletal diversity in plants. Phytochemistry. 2009;70:162137.
Dicke M. Behavioural and community ecology of plants that cry for help. Plant Cell Environ.
2009;32:65465.
175
Dyer LA, Dodson CD, Stireman JO, et al. Synergistic effects of three Piper amides on generalist
and specialist herbivores. J Chem Ecol. 2003;29:2499514.
Fatouros NE, van Loon JJA, Hordijk KA, et al. Herbivore-induced plant volatiles mediate in-flight
host discrimination by parasitoids. J Chem Ecol. 2005;31:203347.
Fine PVA, Mesones I, Coley PD. Herbivores promote habitat specialization by trees in Amazonian
forests. Science. 2004;305:6635.
Fine PVA, Miller ZJ, Mesones I, et al. The growth-defense trade-off and habitat specialization by
plants in Amazonian forests. Ecology. 2006;87:S15062.
Futuyma DJ, Agrawal AA. Macroevolution and the biological diversity of plants and herbivores.
Proc Natl Acad Sci. 2009;106:1805461.
Futuyma DJ, Mitter C. Insect-plant interactions: the evolution of component communities. Philos
Trans R Soc Lond B Biol Sci. 1996;351:13616.
Gerber E, Hinz HL, Blossey B. Interaction of specialist root and shoot herbivores of Alliaria
petiolata and their impact on plant performance and reproduction. Ecol Entomol.
2007;32:35765.
Gershenzon J, Dudareva N. The function of terpene natural products in the natural world. Nat
Chem Biol. 2007;3:40814.
Gouinguene SP, Turlings TCJ. The effects of abiotic factors on induced volatile emissions in corn
plants. Plant Physiol. 2002;129:1296307.
Hakes AS, Cronin JT. Environmental heterogeneity and spatiotemporal variability in plant defense
traits. Oikos. 2011;120:45262.
Halitschke R, Stenberg JA, Kessler D, et al. Shared signals -alarm calls from plants increase
apparency to herbivores and their enemies in nature. Ecol Lett. 2008;11:2434.
Hopkins RJ, van Dam NM, van Loon JJA. Role of glucosinolates in insect-plant relationships and
multitrophic interactions. Annu Rev Entomol. 2009;54:5783.
Huang T, Jander G, de Vos M. Non-protein amino acids in plant defense against insect herbivores:
representative cases and opportunities for further functional analysis. Phytochemistry.
2011;72:15317.
Ibrahim MA, Nissinen A, Holopainen JK. Response of Plutella xylostella and its parasitoid
Cotesia plutellae to volatile compounds. J Chem Ecol. 2005;31:196984.
Irwin RE, Adler LS. Correlations among traits associated with herbivore resistance and pollination: implications for pollination and nectar robbing in a distylous plant. Am J Bot.
2006;93:6472.
Janz N, Nylin S. The oscillation hypothesis of host-plant range and speciation. In: Tilmon KJ,
editor. Specialization, speciation, and radiation: the evolutionary biology of herbivorous
insects. Berkeley: University of California Press; 2008. p. 20315.
Johnson MTJ, Agrawal AA, Maron JL, Salminen J. Heritability, covariation and natural selection
on 24 traits of common evening primrose (Oenothera biennis) from a field experiment. J Evol
Biol. 2009;22:1295307.
Kaplan I, Halitschke R, Kessler A, et al. Physiological integration of roots and shoots in plant
defense strategies links above-and belowground herbivory. Ecol Lett. 2008;11:84151.
Kessler A, Baldwin IT. Defensive function of herbivore-induced plant volatile emissions in nature.
Science. 2001;291:21414.
Koornneef A, Pieterse CMJ. Cross talk in defense signaling. Plant Physiol. 2008;146:83944.
Koricheva J. Interpreting phenotypic variation in plant allelochemistry: problems with the use of
concentrations. Oecologia. 1999;119:46773.
Kostenko O, Bezemer TM. Intraspecific variation in plant size, secondary plant compounds, herbivory and parasitoid assemblages during secondary succession. Basic Appl Ecol. 2013;14:33746.
Kursar TA, Coley PD. Convergence in defense syndromes of young leaves in tropical rainforests.
Biochem Syst Ecol. 2003;31:92949.
Kursar TA, Dexter KG, Lokvam J, et al. The evolution of antiherbivore defenses and their
contribution to species coexistence in the tropical tree genus Inga. Proc Natl Acad Sci.
2009;106:180738.
176
A.M. Trowbridge
Plant-Microbe Interactions
David A. Lipson and Scott T. Kelley
Contents
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Mutualisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
N-Fixing Mutualisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Mycorrhizae . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Leaf Endophytes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Host Controls Over Mutualisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Plant Growth-Promoting Rhizobacteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Plant-Microbe Signaling in the Rhizosphere . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Nutrient Relations in the Rhizosphere . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
The Rhizosphere Effect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Plant-Microbe Competition for N . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Impact of Plants on Soil Microbial Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
PMI Effects on the Soil Matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Impacts of Plants on Microbial Diversity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Culture-Independent Characterization of Microbial Diversity . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Microbial Diversity of the Phyllosphere . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Microbial Diversity of the Rhizosphere . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Impacts of Microbes on Plant Diversity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
The Role of Plant-Microbe Interactions in Global Change . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Future Directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
178
179
180
185
188
189
190
191
191
191
192
193
195
195
195
196
198
199
200
202
203
Abstract
177
178
Introduction
Plant-microbe interactions (PMI) are central to the functioning of terrestrial ecosystems. No model of plant biology is complete without taking into account their
associated microbes, just as soil microbes cannot be understood without considering the plants that shape their habitat. In fact, given the origin of chloroplasts and
mitochondria from endosymbiotic bacteria, it could be argued that PMI are inherent
to the very biology of plant cells. PMI form a continuum, ranging from highly
coevolved, species-specific mutualisms tightly associated with plant tissues, to the
more variable and general communities of microbes in the soil, which produce
strong feedbacks that drive plant growth and, in turn, are largely controlled by plant
chemistry and microclimate. Plant-microbe mutualisms show an extraordinarily
intricate signaling/gene expression network between host and symbiont, but even
some of the more general PMI are mediated by surprisingly complex and intimate
interactions. PMI can shape both the plant and microbial communities and provide
strong feedbacks in important global processes, such as biological invasions and
climate change. Important gaps remain in the current understanding of PMI, but
methodologies and research are advancing rapidly that may address some of
these gaps.
Plant-Microbe Interactions
179
Fig. 1 A summary of the PMI and their roles in the environment discussed in this chapter
This chapter first describes the nature of the PMI at the individual plant-microbe
level, working from specific to more general associations. It then considers largerscale implications of these interactions for plant and microbial communities,
ecosystems, and global change. The chapter concludes with an assessment of the
current gaps in knowledge and how newly developed tools could help fill those
gaps. Highlights of the topics discussed here are depicted in Fig. 1. We do not deal
explicitly with plant pathogens. However, mutualistic associations such as mycorrhizae can span the mutualism-parasitism continuum, and so parasitism is considered briefly within this context. There is also some mention of pathogenic microbes
in the discussion of microbial impacts on plant diversity. Similarly, while this
chapter deals mainly with positive associations, plant-microbe competition is also
considered, as it is an inherent factor in the functioning of the rhizosphere.
Mutualisms
Mutualisms are differentiated from other positive associations in that they are
generally essential (at least in practical terms) for the survival of one or both
partners, species-specific and show an especially high degree of coevolution
between the partners. The most widely studied plant-microbe mutualisms are
those between leguminous plants and nitrogen (N)-fixing bacteria (collectively
180
N-Fixing Mutualisms
Dinitrogen gas (N2) makes up 80 % of the earths atmosphere, yet N is the most
commonly limiting nutrient for plant growth in most terrestrial ecosystems (with
the exception of the tropics, where phosphorus (P) is more commonly limiting).
One of the main reasons for this apparent paradox is that because of the highly
Stable NN triple bond, enzymatic fixation of N2 into a biologically useable form
is an energetically expensive process, requiring about 16 mol ATP per mole N
fixed. While only prokaryotes (Bacteria and Archaea) are known to carry out this
process, a great variety of plants maintain mutualisms with N-fixing bacteria.
Rhizobia-Legume Mutualism
Most globally important is the legume-rhizobia mutualism. Leguminous plants
(in the bean family, Fabaceae, formerly Leguminosae) form mutualisms with
nodule-forming, N-fixing bacteria, known collectively as rhizobia. These bacteria
are generally Alphaproteobacteria from closely related genera such as Rhizobium,
Azorhizobium, Bradyrhizobium, Mesorhizobium, and Sinorhizobium. More
recently, members of the Betaproteobacteria (such as Burkholderia) have been
found to form similar associations with Mimosa, these sometimes referred to as
beta-rhizobia. There are also reports of Gammaproteobacteria capable of producing nodules and fixing N in legumes (Shiraishi et al. 2010). Rhizobia typically form
nodules on roots, though may also form these structures on stems, as is the case for
the mutualism between the tropical tree, Sesbania rostrata, and its partner,
Azorhizobium spp. The nod genes required for nodulation and the nif genes required
for N fixation are often found on a Sym plasmid or other mobile genetic element. In
particular, the nod genes appear to have been horizontally transferred among the
various nodulating bacteria of the -, -, and even -Proteobacteria (MassonBoivin et al. 2009; Shiraishi et al. 2010).
The infection and nodulation process involves a complex interplay between
bacteria and host. The signaling and genetics have been reviewed in great detail,
especially for the Sinorhizobium-Medicago and Mesorhizobium-Lotus systems
(Oldroyd et al. 2011). The general picture is described here, though these relationships are impressively diverse and likely include many exceptions (Masson-Boivin
et al. 2009). Roots produce flavonoids that attract Rhizobia from the surrounding
soil (where they can survive independently of plants, but will not generally fix N).
Rhizobia bind specifically to lectins (proteins that bind carbohydrates) on the
surface of root hairs via sugar residues on the bacterial surface. Bacterial cell
surface polysaccharides also appear to play key roles in avoiding the hosts defense
response. The bacteria invade the root hair, producing polysaccharide-degrading
enzymes such as polygalacturonase or cellulose to soften the root hair cell wall.
Plant-Microbe Interactions
181
182
Even after nodules are established, there is further complex interplay between
the bacteroid and host. A bacteroid together with its surrounding membrane is
called a symbiosome. Symbiosomes resemble organelles, such as mitochondria or
chloroplasts, in that they are highly dependent on their hosts. Amino acid synthesis
is shut down in bacteroids of some rhizobia species, who instead rely on their hosts
for amino acids (Oldroyd et al. 2011). Similarly, rhizobia require homocitrate from
the host to produce the iron-molybdenum (FeMo) cofactor of nitrogenase, the
primary enzyme of N fixation. The plant host is also primarily responsible for
regulating O2, which is inhibitory to nitrogenase. Plants control the overall O2
permeability of the nodule through a barrier in the nodule cortex and produce the
O2-binding protein, leghemoglobin, reducing free O2 concentrations to the
nanomolar range while facilitating O2 diffusion to the rapidly respiring bacteroids.
The plant supplies energy to the bacteroids in the form of organic acids such as
malate, succinate, and fumarate. The primary product of N fixation, NH4+, is
transported across the peribacteroid membrane, assimilated into glutamate and
glutamine in plant cytosol and exported from the nodule as other N-rich amino
acids such as arginine. An extraordinary amount of detailed information is known
about the few model legume-rhizobia systems mentioned above. However, nature is
full of surprising variation on these themes. Striking examples include the discovery of methylotrophy in the legume-nodulating -Proteobacterium,
Methylobacterium nodulans, and photosynthesis in a group of Bradyrhizobium
spp. that nodulate legumes of the Aeschynomene genus. In both cases, the unexpected energy-generating metabolism by the bacterial endosymbionts (recycling of
plant-produced methanol and photosynthesis in the nodule, respectively) appear to
contribute to the efficiency of the mutualisms (Masson-Boivin et al. 2009).
Another variation is the association of rhizobia with Parasponia, a
non-leguminous genus of tropical tree. These associations appear to be less efficient
and sophisticated than those in legumes (Santi et al. 2013). In terms of host plant
phylogeny and nodule morphology, these mutualisms are similar to actinorhizal
associations, which form with a very different group of bacteria.
Actinorhizal Associations
Actinorhizal N-fixing associations have not been as thoroughly studied as those of
legumes and rhizobia, but they are globally important, contributing possibly about
25 % of terrestrial nitrogen fixation worldwide (Pawlowski and Newton 2008).
Actinorhizal associations form between various plants (generally woody trees or
shrubs) in the Fagales, Cucurbitales, and Rosales and the filamentous
Actinobacterium, Frankia. Actinorhizal nodules have a coral-like morphology,
with multiple lobes. The actinorhizal plants appear to form a single N-fixing
clade of angiosperms with the legumes. Actinorhizal nodules are morphologically
similar to those of legumes, and deeper parallels may also exist. For example, the
receptor-like kinase SymRK is required for nodulation in both legumes in the
actinorhizal tree, Casuarina glauca (Masson-Boivin et al. 2009). Actinorhizal
plants fill similar niches as legumes, for example, in colonizing N-poor soils in
early succession. In cooler regions, actinorhizal plants are often the dominant N
Plant-Microbe Interactions
183
fixers among trees and shrubs, whereas leguminous trees generally fill these niches
in the tropics. However, the Myrica faya invasion of Hawaii (discussed below) is
clearly an important exception to this. The specificity between host and bacterial
endosymbiont varies. Some species of Myrica, such as M. pensylvanica and
M. californica, can host a broad diversity of Frankia, whereas as M. gale has
greater specificity. Myrica, Alnus, Dryas, and Elaeagnus species often have the
ability to form nodules outside their native ranges.
There are a number of parallels between actinorhizal and legume-rhizobia
mutualisms. For example, as in legumes, there are two infection strategies in
actinorhizal symbioses: root hair infection and intercellular colonization (Fig. 3).
The mechanisms of plant-microbe communication in the actinorhizal nodulation
process are known in far less detail than for the legume-rhizobia system, but there
are likely several parallels. Frankia produces root hair deformation factor (Had),
possibly similar to rhizobial nodulation factors, except it does not cause cell
division in the root cortex. Flavonoids may be involved, both in stimulatory and
inhibitory roles, but these are not clearly characterized. There is also some evidence
that lectins might be involved in bacterial binding at the plant surface.
In contrast to the root hair infection process in legumes (in which rhizobia stay in
an extracellular infection thread until reaching the root cortex), one pathway of
infection by Frankia is to enter the cytoplasm of a deformed root hair cell (Fig. 3a).
An infection thread-like structure is formed by growth of host plasma membrane
and cell wall material around the invading Frankia filament. In response to root hair
184
infection by Frankia, root cortical cells below the infected root hair start to divide,
expand, and eventually become infected by Frankia, forming the prenodule. As the
prenodule matures, Frankia differentiate into vesicles (analogous to bacteroids in
rhizobia) and nif genes are expressed. Meanwhile, in the root pericycle, a nodule
primordium is initiated, which expands, becomes infected intracellularly by
Frankia from the prenodule, and develops into a mature nodule lobe. Alternatively,
Frankia can invade by growing between epidermal and cortical cells, in an
expanded intercellular zone created by the thickening of host cell walls (Fig. 3b).
Frankia filaments then penetrate the cytoplasm of cortical cells of the nodule-lobe
primordium, which develops into a nodule lobe. In both cases, actinorhizal nodules
are essentially modified lateral roots, unlike the case in legumes.
In actinorhizal mutualisms, the host plant generally plays a smaller role in
nodule O2 regulation than in legumes. Frankia can fix N under aerobic conditions,
in part due to the thick walls and rapid respiration rates of its vesicles. Actinorhizal
nodules generally lack the dense cell layers that restrict O2 in legume nodules, but
some host plants form nodules with thick, lignified cell walls and high levels of
hemoglobin (of either plant or bacterial origin), and in which the Frankia do not
form vesicles. Nodule roots represent an interesting variation in nodule O2
relations: these are produced by some actinorhizal plants (such as Myrica spp.) in
wet soils, growing upwards above the water table to conduct O2 to submerged
nodules through porous (aerenchymous) tissues.
In contrast to rhizobia bacteroids in legume nodules, Frankia assimilates the
NH4+ produced in N fixation and instead exports N to host cells in the form of
amino acids such as arginine (Berry et al. 2011). Actinorhizal mutualisms are not
nearly as genetically well characterized as in legumes. While this knowledge base is
growing and a number of symbiotic genes have been identified, the precise roles for
these genes in symbiosis are still being elucidated.
Plant-Cyanobacterial Mutualisms
Heterocystous cyanobacteria are filamentous photosynthetic bacteria with specialized cells (heterocysts) where N fixation takes place. Heterocystous cyanobacteria
form N-fixing mutualisms with bryophytes, the water-fern Azolla (Pteridophyta),
cycads (Gymnosperms), and the flowering plant Gunnera (an angiosperm). Cycads
are among the more ancient lineages of extant vascular plants and arose much
earlier than the nodule-forming angiosperms. Their N-fixing mutualisms with
cyanobacteria appear to be far less sophisticated than those in nodules. Cycads,
when infected by Nostoc spp., produce coralloid roots in which the cyanobacteria
are housed in a mucilaginous extracellular space. In terms of this mutualism, more
is known about gene expression in the cyanobacterial partner, which has slower cell
division, increased cell volume, altered intracellular structures, and increased
frequency of heterocysts compared to the free-living state. However, the
cyanobacteria are far more independent of their hosts and less physiologically
altered than in the previously mentioned mutualisms. The exact C source provided
by the plant is currently not certain but may be simple sugars. Fixed N is assimilated
into amino acids (glutamine or citrulline) within the heterocysts and transferred to
Plant-Microbe Interactions
185
the plant. The cyanobacteria are solely responsible for protecting their nitrogenase
enzymes from O2. This is done by concentrating all N-fixing activity into heterocysts with thick walls and rapid respiration rates (Santi et al. 2013).
The heterocystous cyanobacterium, Nostoc azollae (formerly Anabaena
azollae), grows in cavities on the underside of leaves in the aquatic fern, Azolla.
This may represent the simplest of all plant-bacterial N-fixing associations,
yet some degree of coevolution has occurred. For example, the cyanobiont of
Azolla is transmitted from generation to generation via megasporocarps (structures
Azolla uses for dispersal of its spores), rather than relying on a fresh supply of
cyanobacteria from the environment for each new generation of plant
(Santi et al. 2013).
Mycorrhizae
Mycorrhizal associations form between a wide variety of plant roots and fungi
(Smith and Read 2008). The majority of plant species have some form of mycorrhizae, notable exceptions being the Brassicaceae and Chenopodaceae families.
These two non-mycorrhizal plant families are generally ruderal (weedy) and so
grow best in high-nutrient conditions where mycorrhizae would be of less benefit
(see section Host Controls over Mutualisms). Three broad classes of mycorrhizae
are differentiated by the arrangement of fungal hyphae in or around plant cells:
endomycorrhizae penetrate into the plant cytoplasm, ectomycorrhizae (EM) form a
dense mantle around stunted lateral roots and grow between root cells without
penetrating the cell membrane, and ectendomycorrhizae both penetrate into the
interior of the host cells while also forming a mantle. Despite each category being
quite diverse, these morphologies are fairly well correlated with their ecological
roles. These associations mainly provide benefit by effectively extending the root
surface for nutrient uptake; but they also may offer the host plant some
protection from pathogens or other stresses such as heavy metal toxicity.
The most widespread and well-studied type is the arbuscular mycorrhizae (AM),
a form of endomycorrhizae.
Arbuscular Mycorrhizae
AM are the most common mycorrhizae, forming endomycorrhizal associations
with about two thirds of all plant species and about 80 % of angiosperms. However,
AM are also found among gymnosperms, bryophytes, and ferns. They are particularly common among herbaceous species, and so from an ecosystem perspective,
AM is the dominant type of mycorrhizal relationship in grasslands. The fungal
partners are now placed in the Glomeromycota phylum. These fungi are reliant on
their hosts and are therefore considered obligate biotrophs. As such they do not live
independently as saprotrophs (decayers of dead organic matter) in soil, but exist in a
dormant form until they encounter a compatible host root. As a result, no AM
mycobiont exists as a pure culture, though co-cultures with plant root tissue have
been maintained.
186
Because these fungi are not effective saprotrophs, they are less able to access N
that is covalently bound to complex soil organic matter. P is bound to organic
matter through ester bonds that require a narrower class of enzymes to cleave, and
so the primary nutritional role of AM is to acquire phosphorus (P) for their hosts.
AM are named for the highly branched, treelike structures (arbuscules) they form
within plant cortical cells. These structures are the primary site of nutrient exchange
between the fungus and plant (the highly branched geometry provides high surface
area for exchange). Vesicles are also found within the roots in some AM, and in
older literature, the term vesicular-arbuscular mycorrhizae (VAM) is frequently
found. These structures appear to play a storage/dormancy role in the fungus and
are capable of infecting new roots. To survive in the soil between hosts, AM fungi
produce spores, such as the large gigaspores of Gigaspora spp., which can reach
about 0.5 mm in diameter.
The AM infection process has been worked out in detail for Medicago truncatula
(Bonfante and Genre 2010). There are a multitude of signals between plant and
fungus, in which each senses and responds to the other. The fungus senses the
presence of roots through root exudates and CO2 from root respiration. These
signals stimulate spore germination. In fact, spores can be germinated in the lab
under elevated CO2 but will abort without the presence of a compatible host. Fungal
hyphae are stimulated to become highly branched when in close proximity to a host
root. The exact signal for this response is unknown at this time, but the response is
produced most strongly in P-starved plants. The plant senses the approaching
fungus even before physical contact with the roots, through an unresolved soluble
signaling molecule. The fungus must avoid triggering the defense response of the
host plant. This may be done by altering chitin in the fungal cell walls, degrading
the plant-produced defense signals, or producing defense-suppressing compounds.
The plant undergoes systemic changes (found in the entire plant rather than just the
local infected area) in response to AM infection. These include expression of
P-starvation genes and lateral root formation, both serving to increase the efficiency
of infection. Additionally, infection induces cell-specific gene expression in roots,
such as cellulase, chitinase, and P uptake. The cellulase enzymes presumably act to
soften the plant cell wall to allow intracellular penetration, whereas the chitinase
could be part of the plants general defense response.
Plant-Microbe Interactions
187
mats of hyphae (mantles) around stunted lateral roots, called club roots, due to their
club-like appearance. The fungal hyphae penetrate between plant cortical cells,
forming a network referred to as the Hartig net. EM greatly extend the length and
surface area of the rooting system. Because fungal hyphae have a smaller diameter
than plant cells, it is much cheaper for a plant to allocate C to its EM fungus than to
produce the equivalent amount of root length or surface area. EM relationships
range greatly in the specificity between host and mycobiont. A single host plant
may be infected simultaneously by a high diversity of EM fungi, while on the other
extreme, some plant species have very specific requirements for infection and their
range may be restricted by the presence of compatible EM fungi in the soil.
Ectendomycorrhizae form a mantle and Hartig net like those of EM but also
penetrate into plant epidermal and cortical cells. These are formed by Ascomycetes
on species of Pinus and Larix. Interestingly, the same fungal species can form ecto-,
ectendo-, or ericoid mycorrhizae, depending on the host plant, illustrating how the
plant controls the morphology of these structures. Arbutoid mycorrhizae, formed in
Arbutus species of the Ericales, also form a mantle, Hartig net, and intracellular
structures, but are distinguished from ectendomycorrhizae in that they only infect
epidermal cells.
Ericoid Mycorrhizae
Ericoid associations are found among the Ericaceae plant family, including many
wetland species, and so these are the predominant mycorrhizal type in wetlands.
The fungal partners are Ascomycetes and Deuteromycetes. As complex organic
matter accumulates in wetland soils, ericoid fungi appear to be adapted to access N
from highly complex organic molecules and so can allow their hosts access to forms
of organic N not generally available to other plants. Ericoid mycorrhizae form
intracellular coils, which function analogously to arbuscules in plant-fungus nutrient exchange.
Orchid and Monotropoid Mycorrhizae
These two mycorrhizal types are grouped together here because both include
non-photosynthetic plants that use fungi to access organic carbon from other plants
or decaying organic matter. Plants from the Orchidaceae (orchids) form obligate
mycorrhizal relationships with Basidiomycete fungi (and a few Ascomycetes).
Orchids produce very small seeds without major storage reserves. As a result,
they rely on mycorrhizae for seed germination and early establishment of seedlings.
Some orchids are non-photosynthetic and mycoheterotrophic, meaning they rely on
these associations throughout their life, while others are mixotrophic, gaining C
both from photosynthesis and mycorrhizal fungi. This relationship is unique among
mycorrhizae in that the plant is reliant on organic C from the fungus, whereas
typically the plant provides C to the mycobiont in exchange for nutrients. These
relationships gain C by parasitizing ectomycorrhizal networks of other plant species
or by the saprotrophic activity of the fungal partner. One noteworthy example of the
latter form of relationship is that between some mycoheterotrophic orchids (such as
Galeola and Gastrodia) and Armillaria mellea, a wood-degrading root pathogen.
188
The orchid produces an antifungal protein, gastrodianin, which may play a role in
preventing root degradation by the fungus (Baumgartner et al. 2011). Armillaria is
bioluminescent, generating light using the enzyme, luciferase. It is possible that the
bioluminescence attracts nocturnal animals for fungal spore dispersal.
Many orchids form associations with the so-called Rhizoctonia complex,
actually comprising three groups within the Agaricomycetes: the Sebacinales,
Ceratobasidiaceae, and Tulasnellaceae. Rhizoctonia-type associations are
endomycorrhizal: fungi form coils (pelotons) between the cell wall and membrane
of root cortical cells. These pelotons are eventually digested by the plant. It is
still uncertain to what extent the fungi in these associations benefit from the
relationships, and the orchids have often been viewed as parasitizing the fungi.
However, this view may be changing as evidence for the mutualism of these
relationships emerges (Dearnaley et al. 2012).
Monotropoid mycorrhizae are also formed by non-photosynthetic, parasitic
plants. Like some of the orchid mycorrhizae, Monotropoideae species (Ericales)
tap into EM networks to access sugars and nutrients. These appear to be exploitative
mycorrhizae rather than mutualisms, in that there is no evidence that the fungi
benefit. As in classic ectendomycorrhizal structures, a mantle, Hartig net, and
intracellular hyphae are formed. The exchange of nutrients presumably occurs in
the fungal pegs that penetrate into epidermal root cells.
Leaf Endophytes
Mutualisms between microbes and plant roots have received the most attention, but
there are a number of important and fascinating mutualistic associations between
microbes and leaves. For example, endophytic associations between grass and fungi
can protect the host from herbivory, disease, and drought stress and can stimulate
Plant-Microbe Interactions
189
root growth (Saikkonen et al. 2013). Epichloe species (Ascomycetes) are common
leaf endophytes in grasses. Like mycorrhizal relationships, leaf endophytic relationships can range from mutualistic to parasitic. In the more mutualistic instances,
the fungus is transmitted to new generations of host plants through seeds. In these
symbioses, the fungus does not penetrate cell walls, often colonizing vascular
tissues. The growth of fungal hyphae and plant tissues are well coordinated, with
hyphal growth ceasing once leaf elongation is complete. The fungi protect plants
from insect herbivores through production of alkaloids, such as peramine and
loline. Some strains produce indolediterpenes and ergot alkaloids, which are also
effective against vertebrate herbivores (the latter, including lysergic acid amine, is
often responsible for poisoning livestock). Endophytic fungi can provide protection
against root-feeding nematodes, despite their absence in roots. This might be
caused by translocation of toxins synthesized by the fungus, induction of plant
defenses, or morphological changes in the roots of the host plant. Stimulation of
root growth by endophytes may also be responsible for increased stress resistance in
the host.
Leaf surfaces also support thriving communities of bacteria, including numerous
beneficial species (see Microbial Diversity of the Phyllosphere section below).
Some, such as Sphingobacterium spp., provide protection from leaf pathogens
(Vorholt 2012). In the nonvascular realm, Sphagnum spp. living in methanogenic
wetland ecosystems host CH4-oxidizing bacteria that convert CH4 to CO2, which
the host plant uses for photosynthesis (Raghoebarsing et al. 2005).
190
Plant-Microbe Interactions
191
nodulation in legumes. Some PGPR produce plant hormones that can have a
positive impact on plant growth. Finally, the presence of benign bacteria on the
root surface can have a probiotic effect, protecting the root from opportunistic
pathogens by keeping this niche occupied (Santi et al. 2013).
192
result from the dynamics of root growth in the soil. Exudates are produced maximally near the growing root tip, and so as the root moves through the soil, it creates
a dynamic boom-bust pattern in its wake, causing previously stimulated microbes
to release their N upon starvation. These processes are illustrated in Fig. 4.
Plant-Microbe Interactions
193
roots (driven by transpiration at the leaf surface). Also the higher mobility of nitrate
allows plant roots to create a larger depletion zone around the roots, leading to a
stronger concentration gradient and more rapid diffusion to the root surface.
Because microbes are supposed to outcompete plant roots during nutrient uptake
and because soil microbial growth should generally be limited by either C or N, it
was initially surprising to learn that some plants acquire significant amounts of their
N budget from the uptake of organic forms of N. In particular, amino acid uptake
has proven to be widespread among plants, even when in a non-mycorrhizal state.
In at least one ecosystem, plants and microbes appear to have different preferences
for amino acids: in a study of the alpine sedge, Kobresia myosuroides, and alpine
soil microbes, the smaller, more rapidly diffusing amino acid, glycine, was preferable to plants, whereas the energy-rich and metabolically central amino acid,
glutamate, was taken up more rapidly by microbes (Lipson et al. 1999). These
complementary preferences are consistent with the fact that plants are autotrophic
and therefore limited by mineral nutrients such as N rather than by C, whereas most
soil microbes are heterotrophic and therefore are generally limited by organic C.
When considering short-term direct competition between plants and microbes,
the best predictor of the outcome is probably root and mycorrhizal surface area.
However, it is important to keep in mind the fact that while roots and soil microbes
exist in close proximity, they do not share the same niche, and so plant roots do not
need to outcompete microbes in general for some limiting nutrient such as N. Plants
tend to grow and senesce on a timescale of months to years, whereas soil microbial
biomass turns over many times per year. Also, plants are generally N limited, while
soil microbes are more commonly energy limited. Therefore, in the short-term,
microbes will outcompete plant roots for available N, but in the longer term, plants
gain access to N from the turnover of microbial biomass. In turn, plant tissues
senesce and are decayed by C-hungry microbes. This relationship between plants
and microbes can help retain N in ecosystems: microbes immobilize nutrients when
plants are inactive, preventing gaseous and hydrological losses, and then act as a N
source for plants during the growing season.
194
Plant-Microbe Interactions
195
from saturated soil layers. Mosses, such as Sphagnum, tend to create more waterlogged, anoxic conditions because of their tremendous water holding capacity. The
effect of Sphagnum on soil water content is also an example of how plants can alter
the soil microclimate. These mosses can also act as efficient insulators of soil,
regulating thaw depth in arctic tundra soils. In most ecosystems, either temperature
or soil water content will limit microbial activity at some point. The plant community affects both of these variables through shading, sheltering, and transpiration.
These effects depend on plant community characteristics such as canopy structure,
growth rate, and root:shoot allocation patterns.
196
was less than 1 % of the existing microbial diversity, and often far less. The
development of what is now known as culture-independent molecular techniques
revolutionized the investigation of environmental microbiology and radically
altered our understanding of microbial diversity in countless environments, including those associated with plants. Instead of determining microbial species by
growing them in liquid or solid media before chemical or morphological analysis,
culture-independent methods directly analyze the genetic information in the
microbes, typically the DNA. Also, unlike culturing methods that focus on one
species at a time, culture-independent molecular methods can simultaneously investigate all the members of a particular community using the information of the genetic
sequences to determine the types of microorganisms present in a sample.
The most common gene targeted for this type of analysis is the small-subunit
ribosomal RNA (rRNA) gene, known as the 16S rRNA in Bacteria and Archaea.
(In Eukarya it is known as the 18S rRNA because the RNA is significantly larger in
eukaryotes). Small-subunit rRNA gene sequences are effective genetic markers for
culture-independent microbial studies for a number of reasons. First, this gene
sequence is found in all forms of cellular life: Bacterial, Archaeal, and Eukaryal.
Second, there exists a large and rapidly growing database of rRNA gene sequences
from both cultured and uncultured microbes, allowing ready species identification
and phylogenetic analyses. Third, when comparing the sequences of 16S rRNA
genes among organisms, it was found to have both highly conserved regions and
highly variable regions of sequence. For example, some regions of the sequence
were exactly the same between extremely diverse organisms, such as all of the
Bacteria or between E. coli and humans, while other regions were so variable one
can detect sequence differences between different closely related species of
microbes. The conserved regions were critical for designing PCR primers that
could amplify this gene from, for example, all the bacterial species in a soil sample,
while the variable regions were important for telling the species apart. Recently,
these methods have been combined with high-throughput sequencing approaches,
also called next-generation sequencing (NGS). NGS allows the generation of
hundreds of thousands to millions of DNA sequences simultaneously. Furthermore,
using PCR primers labeled with unique barcode sequences at their 50 end, one can
use NGS to describe gene diversity from many environmental samples in a single
sequencing reaction. After the sequencing, computational methods are used to
determine which sequences came from which samples and what organisms are
present in each sample.
Plant-Microbe Interactions
197
grow, providing living space for an astonishing 1 1026 microbial cells! The
morphology of the leaves, including the three-dimensional surface contours and
leaf structures (veins, stomata, trichomes, etc.), the chemical composition of the
surfaces and the local environmental conditions determine to a large degree the
types of microbes that persist and grow on leaves. The top of leaf surfaces is
exposed to direct sunlight and UV radiation, and the waxy cuticle prevents plant
desiccation and helps retain the plants own metabolites. This makes for an oligotrophic (nutrient poor) environment, selecting microbes able to survive and grow in
these stressful conditions. The undersides of leaves are less exposed to light and,
while still covered in a waxy cuticle, tend to retain moisture more readily.
Leaf morphological structures also influence the ability of microbes to colonize
the surface of the leaf. Microbial communities tend to form in clumps, called
aggregates, in the crevices formed at epidermal cell junctions, along the leaf
veins and at the base of trichomes. These aggregate cells can form biofilms by
secreting extracellular polymeric substances to protect from desiccation and other
stresses. The leaf aggregates tend to also be found in the relatively moist surface
depressions. The aggregates contain fungi as well as bacteria, but archaea are rare in
the phyllosphere.
While the presence and abundance of bacteria have long been known, the advent
of culture-independent molecular methods and NGS, in particular, have allowed for
a much deeper appreciation of the true extent of microbial diversity in the
phyllosphere. They are also providing the means for the comprehensive analysis
of microbial communities across hundreds of thousands of samples. This will be
necessary to determine the subtler abiotic and biotic factors affecting phyllosphere
diversity, especially given the enormous environmental variability across environments and even within a single plant (e.g., the microbial diversity of leaf surfaces at
the top versus bottom branches of a redwood tree (Vorholt 2012)).
So, what have culture-independent methods revealed about the diversity of the
phyllosphere? First, the studies done so far have determined that the species
richness tends to be very high and increases as one moves from temperate to
tropical environments. Given that moisture seems to be a limiting factor, this may
be a function of greater rainfall in the tropics and perhaps higher growing temperatures and slower leaf turnover. Second, while species richness is relatively high,
the phyllosphere as a whole is less diverse than in typical soil rhizosphere communities. The phyllosphere is typically more nutrient poor and short-lived than the
rhizosphere. Third, culture-independent analysis of phyllosphere microbial diversity from very different plant species found it was dominated by bacterial species
from a fairly limited range of bacterial phyla. The Proteobacteria, particularly
Alphaproteobacteria families such as Methlyobacteriaceae and Sphingomonadaceae, were dominant, comprising upwards of 70 % of the bacterial species
on leaves. Other common and abundant phyla included the Bacteroidetes and the
Actinobacteria. Of the four different plant species investigated in one study,
researchers found that between 30 and 40 genera of bacteria were consistently
common on leaves, though the proportions of these genera (and certainly the
specific strains or species) varied considerable across the various plant species.
198
In terms of the particular abiotic and biotic factors that determine microbial
community diversity in the phyllosphere beyond the general ones mentioned, there
are still far more questions than answers. What is known is that environmental
factors, such as nitrogen-fertilization, exposure to solar radiation and pollution, as
well as biotic factors such as leaf age, do significantly affect the structure of
microbial communities. Plant genotype also appears to play an important role in
the microbes that persist on leaves. Moreover, overall diversity within a plant
species tends to be consistently lower than between species. For instance, a study
of pine tree phyllosphere microbial diversity found significantly higher microbial
diversity among the phyllosphere of different pine species with overlapping geographic distributions than within the same species.
Plant-Microbe Interactions
199
microbiome differs among different plant species in the same soil types. Moreover,
transplanting a plant species in a different soil can alter that soils microbial
community to resemble the soil from which the plant originated. Different plant
genotypes of Arabidopsis were shown to alter the plant rhizosphere community,
particularly the Alphaproteobacteria and fungal communities. For example, an
Arabidopsis mutant that produced increased phenolic compounds and decreased
sugars had distinct rhizosphere microbiomes compared with wild-type plants.
Interestingly, the effects of different plant genotypes on the rhizosphere can also
alter the levels of pathogenic microorganisms in the soils. For instance, certain
potato cultivars favor higher levels of Pseudomonadales, Streptomycetaceae, and
Micromonosporaceae, all of which are known to control plant pathogens to some
degree. Finally, it appears that plants can alter their recruitment of beneficial
bacteria when under attack by insect herbivores or pathogenic organisms. Several
plant species have been shown to release root compounds that increase beneficial
organisms (e.g., Bacillus subtilis) in the soils when under attack.
200
Plant-Microbe Interactions
201
202
availability will translate into altered N2O fluxes. Changes in plant community
could feedback to alter trace gas production through impacts on soil redox conditions, pH, and chemistry. For example, loss of mosses in the Arctic (say, because of
sensitivity to warmer, drier conditions) could drastically alter soil conditions and
the relative production of CO2 and CH4. Changes in plant communities will lead to
altered microbial communities with different metabolic properties. For example,
changes in fungal:bacterial ratio or overall species composition can affect the
biomass-specific respiration rate (or C use efficiency, CUE), leading to different
amounts of CO2 produced per unit microbial biomass per unit time. The tangled
web shown in Fig. 5 indicates the great complexity of PMI and their implications
for the planet. The magnitudes of these effects are active areas of research.
Future Directions
Numerous gaps still remain in the current understanding of PMI. For example, the
roles of PMI in feedbacks to global change, especially multifactor changes such as
increased CO2 and temperature, are not yet included in climate change models
(Dieleman et al. 2012). Similarly, the mediation of biological invasions by PMI is
an active area of research. There is still much unknown about the genetics of plantmicrobe mutualisms. For example, sequencing the genome of Glomus intraradices
is challenging due to its heterozygosity and lack of a uninucleate stage. And while
rhizobia-legume mutualisms have been studied in great detail, current understanding of non-rhizobial N-fixing symbioses lags behind. And while great progress has
been made towards understanding the factors that control microbial diversity in
soils, a detailed understanding of how plants shape microbial communities in the
phyllosphere and rhizosphere has yet to emerge. Finally, it appears that plants may
be emerging as reservoirs for bacterial pathogens of humans, but this phenomenon
is not yet well understood.
There are numerous new techniques emerging to help answer these lingering
questions. The development of high-throughput sequencing technology and other
molecular techniques is rapidly changing the face of microbial ecology, making the
study of complex microbial communities more tractable. Meanwhile, analytical
techniques are making rapid advancements, allowing sensitive detection of processes at unprecedented spatial and temporal resolution. For example, new technology such as laser and cavity ring-down techniques allow the real-time
measurement of trace gases and their stable isotopes. Novel visualization techniques, such as reporter genes and synchrotron-based methods, are creating new
windows into the rhizosphere (Raab and Lipson 2010). These and other novel
methods allow the quantification and identification of C compounds transported
from plants into the rhizosphere and to root mutualists. Given the urgent nature of
some of the unanswered questions surrounding PMI and the advent of these new
techniques, the next decade should produce some very interesting work in these
areas.
Plant-Microbe Interactions
203
References
Badri DV, Weir TL, Dvd L, Vivanco JM. Rhizosphere chemical dialogues: plantmicrobe
interactions. Curr Opin Biotechnol. 2009;20:64250.
Baumgartner K, Coetzee MPA, Hoffmeister D. Secrets of the subterranean pathosystem of
Armillaria. Mol Plant Pathol. 2011;12:51534.
Berendsen RL, Pieterse CM, Bakker PA. The rhizosphere microbiome and plant health. Trends
Plant Sci. 2012;17:47886.
Berry AM, Mendoza-Herrera A, Guo Y-Y, Hayashi J, Persson T, Barabote R, et al. New perspectives on nodule nitrogen assimilation in actinorhizal symbioses. Funct Plant Biol.
2011;38:64552.
Bever JD, Platt TG, Morton ER. Microbial population and community dynamics on plant roots and
their feedbacks on plant communities. Annu Rev Microbiol. 2012;66:26583.
Bonfante P, Genre A. Mechanisms underlying beneficial plant fungus interactions in mycorrhizal symbiosis. Nat Commun. 2010;1:48.
Cesco S, Mimmo T, Tonon G, Tomasi N, Pinton R, Terzano R, et al. Plant-borne flavonoids
released into the rhizosphere: impact on soil bio-activities related to plant nutrition. A review.
Biol Fertil Soils. 2012;48:12349.
Cipollini D, Rigsby CM, Barto EK. Microbes as targets and mediators of allelopathy in plants. J
Chem Ecol. 2012;38:71427.
Dearnaley JDW, Martos F, Selosse M-A. Orchid mycorrhizas: molecular ecology, physiology,
evolution and conservation aspects. In: Hock B, editor. Fungal associations. 2nd ed. Berlin/
Heidelberg: Springer; 2012. p. 20730.
Dieleman WIJ, Vicca S, Dijkstra FA, Hagedorn F, Hovenden MJ, Larsen KS, et al. Simple additive
effects are rare: a quantitative review of plant biomass and soil process responses to combined
manipulations of CO2 and temperature. Glob Chang Biol. 2012;18:268193.
Ehrenfeld JG. Effects of exotic plant invasions on soil nutrient cycling processes. Ecosystems.
2003;6:50323.
Eviner VT, Chapin FS. Functional matrix: a conceptual framework for predicting multiple plant
effects on ecosystem processes. Annu Rev Ecol Evol Syst. 2003;34:45585.
Hajek T, Ballance S, Limpens J, Zijlstra M, Verhoeven JTA. Cell-wall polysaccharides play an
important role in decay resistance of sphagnum and actively depressed decomposition in vitro.
Biogeochemistry. 2011;103:4557.
Javot H, Penmetsa RV, Breuillin F, Bhattarai KK, Noar RD, Gomez SK, et al. Medicago truncatula
mtpt4 mutants reveal a role for nitrogen in the regulation of arbuscule degeneration in
arbuscular mycorrhizal symbiosis. Plant J. 2011;68:95465.
Kuzyakov Y, Xu X. Competition between roots and microorganisms for nitrogen: mechanisms and
ecological relevance. New Phytologist. 2013;198:65669.
Lipson DA, Raab TK, Schmidt SK, Monson RK. Variation in competitive abilities of plants and
microbes for specific amino acids. Biol Fertil Soils. 1999;29:25761.
Masson-Boivin C, Giraud E, Perret X, Batut J. Establishing nitrogen-fixing symbiosis with
legumes: how many rhizobium recipes? Trends Microbiol. 2009;17:45866.
Mayerhofer MS, Kernaghan G, Harper KA. The effects of fungal root endophytes on plant growth:
a meta-analysis. Mycorrhiza. 2013;23:11928.
Newsham KK. A meta-analysis of plant responses to dark septate root endophytes. New Phytol.
2011;190:78393.
Oldroyd GED, Murray JD, Poole PS, Downie JA. The rules of engagement in the legume-rhizobial
symbiosis. Annu Rev Genet. 2011;45:11944.
Pawlowski K, Newton WE, editors. Nitrogen-fixing actinorhizal symbioses. Dordrecht: Springer;
2008.
Ponge J-F. Plant-soil feedbacks mediated by humus forms: a review. Soil Biol Biochem.
2013;57:104860.
204
Raab TK, Lipson DA. The rhizosphere: a synchrotron-based view of nutrient flow in the root zone.
In: Grafe M, Singh B, editors. Advances in understanding soil environments by application of
synchrotron-based techniques. 1st ed. The Netherlands: Elsevier; 2010.
Raghoebarsing AA, Smolders AJP, Schmid MC, Rijpstra WIC, Wolters-Arts M, Derksen J,
et al. Methanotrophic symbionts provide carbon for photosynthesis in peat bogs. Nature.
2005;436:11536.
Saikkonen K, Gundel PE, Helander M. Chemical ecology mediated by fungal endophytes in
grasses. J Chem Ecol. 2013;39:9628.
Santi C, Bogusz D, Franche C. Biological nitrogen fixation in non-legume plants. Ann Bot.
2013;111:74367.
Shiraishi A, Matsushita N, Hougetsu T. Nodulation in black locust by the Gammaproteobacteria
Pseudomonas sp. and the Betaproteobacteria Burkholderia sp. Syst Appl Microbiol.
2010;33:26974.
Smith SE, Read DJ. Mycorrhizal symbiosis. 3rd ed. New York: Academic; 2008.
van der Putten WH, Klironomos JN, Wardle DA. Microbial ecology of biological invasions. ISME
J. 2007;1:2837.
Vorholt JA. Microbial life in the phyllosphere. Nat Rev Microbiol. 2012;10:82840.
Further Reading
Crespi M, editor. Root genomics and soil interactions. Ames: Wiley-Blackwell; 2013.
Maheshwari DK, editor. Bacteria in agrobiology: stress management. Heidelberg: Springer; 2012.
Pinton R, Varanini Z, Nannipieri P, editors. The rhizosphere: biochemistry and organic substances
at the soil-plant interface. 2nd ed. Boca Raton: CRC Press; 2007.
Contents
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
How Is NPP Measured? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Direct Field Measurement of Aboveground Primary Production . . . . . . . . . . . . . . . . . . . . . . . . . .
Field Approaches for Belowground Primary Production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Remote-Sensing and Modeling Approaches to Terrestrial Primary Production . . . . . . . . . .
Patterns and Controls on Productivity from Global to Local Scales . . . . . . . . . . . . . . . . . . . . . . . . . .
Abiotic Controls on NPP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Temporal Variability in NPP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Disturbance and NPP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Biotic Controls on NPP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Vegetation Structure and NPP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Biodiversity Effects on Productivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Community Change and NPP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Herbivory and NPP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Belowground Productivity: Patterns and Controls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Controls of NPP and the Future . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Future Directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
207
211
212
213
214
216
218
226
229
230
231
232
235
236
237
239
244
244
Abstract
205
206
dominant role in the global carbon cycle with crucial implications for global
climate change. Net primary production (NPP) is the amount of fixed energy
or organic matter left over after the plants have met their own respiratory
needs and represents the amount of energy available to the consumers,
including humans. Across the earths terrestrial biomes, a large range of
NPP is observed with the highest values in tropical forests and wetlands,
intermediate values in temperate forests and grasslands, and lowest in
extremely cold or dry deserts.
Accurate measurement of NPP is challenging despite the simple concept that
it represents the amount of new biomass added to the plants in a given time
period. This is because a significant and highly variable proportion of NPP is
lost from the plants by processes such as herbivory, volatilization, and
carbon flux to the soil. Methods of measuring NPP are diverse, being dependent on the structure and dynamics of the vegetation. For example, harvest
methods in which the aboveground tissues are periodically clipped from
quadrats of known area can be effective for quantifying aboveground NPP
in herbaceous vegetation (e.g., grasslands), whereas in woody vegetation,
the growth of woody tissues must also be measured. Moreover, measurements
of total NPP in terrestrial ecosystems must account for root growth which
can be very challenging. As a result, reliable estimates of total NPP are few.
Plants allocate a large proportion of their fixed energy to their root systems to
fuel additional root growth and to meet their respiratory needs. The proportion of total NPP that goes to belowground NPP ranges from about 25 % to
over 50 % and is higher in ecosystems where the degree of limitation by soil
resources is greater, i.e., dry or nutrient-poor sites. Surprisingly, over 10 % of
NPP is contributed by plants to the soil in the form of rhizosphere carbon flux
including exudation, rhizodeposition, and allocation to mycorrhizal fungi and
other symbionts.
Variation in NPP results from differences among ecosystems in the amount of
photosynthetically active radiation (PAR) reaching the plant canopy, the
amount of that PAR absorbed by the foliage (APAR), the biochemical
efficiency of the plants under optimal environmental conditions, and the
degree to which actual conditions are less than optimal. The APAR depends
in part on the amount of foliage surface area per unit ground area (leaf
area index LAI) which ranges from less than 1 in dry or infertile sites to
over 10 in some resource-rich forests. Large-scale monitoring of estimated
NPP is possible using satellite imagery of reflected solar radiation that can
be converted into vegetation LAI and combined with environmental
measurements that indicate the degree of stress reduction to photosynthetic
activity.
Four principal abiotic factors usually limit the amount of NPP on land light,
water, temperature, and mineral nutrients and all these abiotic factors are
changing rapidly as a result of human activity, with highly uncertain implications for global and local NPP. Commonly, two or more of these abiotic
factors concurrently or sequentially limit NPP, but water deficit is arguably
207
the most widespread single factor constraining global NPP. The effect of
temperature on NPP is most closely related to subfreezing conditions that
limit the length of the growing season in temperate and high-latitude environments. Nitrogen is the most important limiting mineral nutrient in most
ecosystems, although in highly weathered tropical soils where nitrogen-fixing
organisms are abundant, phosphorus may be the most limiting nutrient.
Biotic factors can play a key role in regulating NPP so that human activities
such as vegetation management and introduction of exotic species will exert a
major influence on future patterns of NPP. The effects of biodiversity on
NPP have proven difficult to establish, but experimental tests suggest that loss
of species can reduce NPP particularly if a dominant species is lost or
when species numbers become very low, diminishing complementarity in
resource use by coexisting species. Dramatic shifts in plant community
structure, for example, the ongoing invasion of grassland vegetation by
woody plants, can cause changes in NPP that appear to depend in part on
climate. Consumption of plant tissues by herbivores often can have a negative
effect on NPP, but in many grasslands, compensatory growth responses to
herbivory can result in no reduction in NPP or in some cases even stimulation
of NPP by herbivory.
Temporal variation in NPP results from interannual variation in both environmental and biotic factors as well as pulse disturbance events that can reset
the successional clock. The response of NPP to interannual variation in
rainfall seems to be greatest in semiarid and subhumid environments where
average precipitation is sufficient to sustain highly productive communities
(vs. true deserts). Following natural or human disturbances, forests exhibit a
recurring pattern in which NPP peaks after a few decades of stand development, followed by a decline with age in older stands.
Global environmental changes climate, atmospheric CO2, nitrogen deposition, exotic species introductions, etc. are certain to exert a major influence
on global NPP in the future, but the outcomes are highly uncertain because of
the complex ways in which all these changes interact with one another to
influence the vegetation and NPP. For example, CO2 enrichment experiments
indicate that increasing atmospheric CO2 concentration can significantly
stimulate NPP in young forests, but the effect may be transient because of
progressively greater stress by mineral nutrients unless high N deposition
overcomes this limitation.
Introduction
All heterotrophic organisms, from the poles to the tropics, rely on stable forms of
chemical energy collectively known as organic (carbon containing) matter derived
from biological activity. The energy in virtually all organic matter is ultimately
derived from the sun, and the conversion of solar energy to chemical energy is
accomplished by autotrophs, primarily green plants in terrestrial ecosystems.
208
These autotrophs are incredibly diverse in size (<mm to >100 m in height) and life
span (a few weeks to thousands of years) and vary widely in their population
densities, depending on the resources available. But they all employ a similar
photosynthetic process composed of photochemical and biochemical pathways
that are highly conserved from an evolutionary perspective. Subtle variations
within the photosynthetic process (i.e., C3, C4, and CAM photosynthetic pathways)
can have important implications for determining the amount and global distribution
of organic matter produced by plants. But the striking similarities among all
autotrophs in the fundamental mechanism by which inorganic CO2 is converted
into organic matter allows us to step back and focus more on the external controls of
organic matter production in terrestrial ecosystems and less on physiological
variations among autotrophs. This ecosystem perspective is essential for
accomplishing the goal of this chapter which is to provide a contemporary
and forward looking overview of patterns and determinants of primary production
( organic matter production) in terrestrial ecosystems.
Our planets global carbon cycle, of which atmospheric CO2 is a key component
with direct impacts on climate, depends fundamentally upon terrestrial plant primary production. Why is this so? It is because terrestrial ecosystems account for
approximately two-thirds of the global estimate of total primary production, despite
covering only a quarter of the earths surface. The oceans contribute the remainder.
Moreover, the annual removal of CO2 from the atmosphere by the photosynthetic
activities of terrestrial plants is about 20 times greater than CO2 emissions to the
atmosphere from fossil fuel burning by humans (Fig. 1). Similarly, CO2 emitted
back to the atmosphere from the respiratory activities of plants is about 10 times
that of fossil fuel emissions. Finally, estimates of the amount of carbon stored in
terrestrial plants are almost 100 times greater than annual emissions from fossil fuel
burning. Indeed, carbon stored in terrestrial plant biomass is equivalent to about
75 % of the carbon found in the atmosphere. The fact that the amount of carbon
transferred in and out of the atmosphere by plants is an order of magnitude greater
than fossil fuel inputs points to the importance of understanding the dynamics and
fate of terrestrial plant primary production. However, the relatively small size of
anthropogenic sources of carbon to the atmosphere should not belie their importance. Such emissions have been the dominant cause of the 25 % increase in
atmospheric CO2 levels directly measured in the last 50 years (Fig. 1). Evidence
is overwhelming that a consequence of this alteration to the composition of earths
atmosphere will be global warming, an intensification of the global hydrological
cycle, and an increase in the number and severity of climatic extremes and all of
these climatic changes will affect plant processes and future levels of primary
production. Thus, in order to understand ecological patterns and processes now
and in the future, the determinants of primary production across the wide range of
earths terrestrial ecosystems must be understood, from deserts to tropical forests.
Our current understanding of primary production in terrestrial ecosystems is a
product of literally thousands of studies conducted during the last 100+ years, but
before considering any synthesis of this knowledge, some terms and concepts need
to be defined. The total amount of energy fixed (as CO2 into organic matter) by
209
Fig. 1 Simplified depiction of the global carbon cycle with the central role of processes directly
related to production by plants in terrestrial ecosystems highlighted. Dashed lines from plant
carbon to soil carbon boxes indicate that while plant biomass is the source for most soil carbon,
other processes (not shown) determine how much carbon flows from plants to the soil carbon pool.
Units are arbitrary and relative to simplify comparisons
plants per unit ground area per unit time is termed gross primary production
(GPP). This is the sum of all energy fixed by the autotrophs in the ecosystem. Net
primary production (NPP) is the amount of energy left over after autotrophs have
met their energetic needs through respiration. Thus, NPP is GPP minus respiration
by primary producers. NPP represents the amount of energy available to consumers
(including humans) in an ecosystem. NPP is typically expressed in units of dry
matter (grams m2 year1) rather than units of energy because of the ease of
quantifying plant mass and the simplicity of converting mass to energy for plant
tissues. As an alternative to units of dry plant matter, grams of carbon also are
commonly used to express NPP. Because C content of plant biomass is typically
between 45 % and 50 %, converting between plant matter and plant carbon is
straightforward. The total mass of plants (per unit area) at any point in time is often
referred to as standing crop or simply as biomass. Many ecologists conceptualize
NPP as the amount of new biomass added in a given period of time; however, a
significant portion of the NPP actually does not appear as new plant tissue but rather
is lost from the plant by such pathways as canopy leaching, volatilization, and
especially rhizosphere carbon flux, including allocation to mycorrhizal symbionts.
Quantifying these components of NPP is very challenging.
210
Table 1 Range of NPP and standing biomass (dry matter) for different biomes types (from
Huston and Wolverton (2009) (Data are from estimates of above- and belowground components
combined and global NPP is based on estimates of the spatial extent of each biome)
Biome
Tropical forest
Temperate forest
Boreal forest
Tropical savanna and
grassland
Temperate grassland
and shrubland
Desert
Tundra
Crops
Wetlands
Standing crop
biomass (Mg/ha)
240388
114268
84128
58
Net primary
production (g/m2/
year)
1,5662,502
1,2501,558
380468
1,0801,282
1426
596786
10.614.0
48
812
46
86
102252
178358
6081,008
2,458
2.87.0
1.02.0
8.213.6
8.3
Terrestrial NPP has also been conceptualized by focusing on the ultimate source
of energy the sun. In this approach, think of the vegetation community as a living
machine whose growth and metabolism are driven by incoming solar radiation. In
this framework, NPP depends upon the efficiency with which the photosynthetically active radiation (PAR) that is absorbed by plant leaves is assimilated into
organic matter accumulating in the vegetation. Variation in NPP is the result of
differences among ecosystems in the amount of PAR reaching the canopy, the
amount of that PAR absorbed by the foliage (APAR), the biochemical conversion
efficiency of the plants under optimal environmental conditions, and the degree to
which actual conditions are less than optimal. Thus,
NPP E APAR
In this framework the conversion efficiency (E dimensionless) would account
for the photochemical efficiency of leaves, energetic costs of growth and maintenance of plant tissues, as well as any environmental stresses, like drought and cold,
that reduce photosynthesis below optimal. The APAR term accounts for variation in
the amount of PAR reaching the top of the plant canopy, as influenced by day
length, cloud cover, etc., as well as the amount of foliage in the plant community
(leaf area index or LAI the leaf surface area per unit ground area) and its
architectural arrangement. The LAI depends in part on the availability of soil
resources (water, mineral nutrients) and ranges from less than one in deserts to
over 10 in some resources-rich forests. Obviously leaves deep in the canopy of such
forests receive only enough PAR for minimal photosynthesis, and the energetic
costs of growing all the plant tissues leaves, stem, roots in these ecosystems set a
limit on the maximum NPP attained by terrestrial vegetation (Table 1).
Finally, an alternative to the plant-focused considerations above is a carbon
balance perspective on primary production. In this framework, GPP is the total
211
amount of CO2 (or carbon) that is fixed or taken up by plants in the ecosystem, ER
(ecosystem respiration) is the amount of CO2 that is lost or emitted from the
ecosystem from the combined metabolic activities of plants and heterotrophs
including decomposers (microbes). Net ecosystem production or NEP is thus
GPPER or the net amount of primary production after losses to respiration by
plants, heterotrophs, and decomposers. NEP is a valuable measure for evaluating the
balance of CO2 between ecosystems and the atmosphere. Ecosystems sequester or
store carbon when NEP is positive, with the length of time (residence time) this
carbon remains in the ecosystem determined by its turnover rate. The turnover rate
is simply the ratio of standing biomass to NPP. Biomass and NPP are mechanistically related to each other, and in general greater NPP will lead to greater standing
biomass in terrestrial ecosystems. However, the relationship between NPP and
biomass is actually more complex. In forests, for example, aboveground biomass
plateaus at intermediate levels of aboveground NPP and may even decline at the
highest levels of productivity (Fig. 2). This is because turnover rates may increase in
high productivity forests limiting additional biomass accumulation. This relationship is further complicated when comparing the NPPbiomass relationship in
different biome types. For example, some forests may have very high standing
biomass but low NPP in part due to high respiration rates in large trees; the residence
time of C stored in such a system is relatively long and turnover is slow. Conversely,
most grasslands have low standing biomass due to consumption by animals or fire,
even with relatively high NPP. Indeed, some wetlands have levels of NPP that can
match tropical forests, but standing biomass is much lower (Table 1).
212
Such knowledge can be critical for interpreting research and making broader
inferences. The accurate measurement of primary production can be very challenging despite the simple concept that it is the amount of new biomass added to the
vegetation in a time interval. The principal difficulty is that not all the new biomass
that was added is retained at the end of a measurement interval (whether a month,
growing season, or year). In most ecosystems a significant proportion of the NPP
can be lost to processes such as herbivory. Moreover, direct measurement of
changes in the biomass of some tissues, like roots, is very difficult, and a substantial
proportion of the belowground production is lost through a variety of rhizosphere
carbon flux processes such as exudation, rhizodeposition, and allocation to mycorrhizal fungi.
213
not accurate (see Biotic Controls on NPP). A common solution to this problem is
using many temporary, movable exclosures that allow estimation of herbivore
consumption and regrowth responses of grazed plants.
Field measurement of ANPP in ecosystems dominated by woody vegetation
presents challenges owing to the large and complex dimensions of the plants. The
principle underlying field approaches is that
ANPP B M
where B equals the annual increment in live tree biomass and M equals the losses
of living tissue to mortality, including litterfall, pruning/herbivory, and tree death.
The reason that M must be added to B to estimate ANPP is clear for the case
where B is zero: if the live biomass is constant from year to year, and losses of live
biomass are occurring, then the plants must have replaced this lost biomass in the
form of new tissue production. Because of the large size of the plants, B is usually
estimated by applying allometric equations that describe the relationship between
an easily measured dimension of the plant (e.g., stem diameter or tree height) and
plant biomass. Such equations have been developed for most common woody plant
species or groups. Next, the annual or multiyear change in diameter can be used to
estimate B for each plant in the sample plot. The largest aboveground loss of
ephemeral tissue contributing to M is fine litterfall which is easily collected using
littertraps. Note that in mature woody vegetation, the amount of leaf litterfall is about
the same as the new foliage production. For some other litterfall components, especially fruits and woody tissues, the amounts can vary a lot from year to year, and
several years of collection will be needed to adequately account for annual variation.
If B is estimated on the basis of multiyear changes in live tree biomass, then the
value of M must also account for trees that died during the measurement interval; this
is usually accomplished with tagged tree inventory, but the plots must be large enough
to overcome the high spatial variability in tree mortality. Finally, the measurement of
ANPP will be incomplete if understory vegetation is ignored and suitable adaptation
of herbaceous and woody vegetation measurements may be needed.
214
single component of BNPP is usually the growth of ephemeral fine roots, defined as
smaller than some arbitrary diameter cutoff (e.g., <1 mm). These fine roots are very
important functionally as well for water and nutrient uptake from the soil. A smaller
fraction of BNPP goes to the growth of long-lived coarse roots. As noted earlier, a
large proportion of BNPP is conveyed to mycorrhizal symbionts or exported from
the roots by passive exudation or active rhizodeposition. Finally, a fraction of
BNPP is lost to root herbivory, but methodological challenges have limited these
measurements to just a handful of studies.
Most estimates of TRA and BNPP in natural vegetation employ a steady-state
assumption for either or both soil C content or fine root biomass, meaning that the
steady-state parameter is neither increasing nor decreasing substantially. Under this
assumption TRA can be estimated as the difference between the annual emission of
CO2-C from the soil (total soil respiration TSR) and annual aboveground
litterfall, both of which can be measured with high accuracy. Thus, reliable
estimates of TRA are available for a variety of global vegetation types. However,
to calculate BNPP from TRA requires accurate measurement of Rr which is
challenging because of the complexity of plant root systems, their highly variable
metabolic activity, as well as the intimate contact between roots and soil and
attendant microbes. Nevertheless, the accurate measurement of TRA has provided
useful insights into patterns of BNPP in relation to biotic and environmental factors
(see below).
Again, the largest component of BNPP is the growth of short-lived fine roots
(FRP). The most reliable way of estimating FRP is combining field measurements
of fine root biomass and indices of root turnover, i.e., the proportion of the fine root
biomass dying and being replaced annually. Under the steady-state assumption, the
fine root turnover coefficient (TC, year1) is the inverse of the average root
lifespan, and FRP can be calculated as the product of TC and average fine root
biomass. The latter is measured by coring the soil and laboriously sorting the live
roots from the soil. Several approaches have been used to estimate fine root TC,
including minirhizotrons with which roots can be viewed growing along the surface
of a transparent tube inserted into the soil and their survivorship monitored through
time (Fig. 3). Measurements of TC based on the decay or dilution of isotopes also
have been achieved, but in all cases, a variety of sources of error and bias must be
overcome, as summarized by Tierney and Fahey (2007).
215
1.0
0.8
0.6
0.4
Cohorts (1996)
0.2
0.0
May
June
September
October
May
September
1996
May
September
1997
Fig. 3 An example of fine root survivorship data from a minirhizotron tube (a transparent tube
inserted into the soil) beneath a northern hardwood forest in northeastern USA. Periodically, a
camera is lowered into the tube and images of roots growing along the surface of the tube are
recorded. By identifying and noting the location of a number of roots at one point in time
(a cohort), the survival or disappearance of these roots can be reassessed at regular intervals
over time. Note in the data above that there were only a few exceptions; fine roots disappear over
time with some cohorts (October 1996) experiencing 100 % mortality in less than a year. Root
survivorship can be used to estimate the turnover coefficient for calculations of fine root production (adapted from Tierney and Fahey 2001)
these approaches allow both high-resolution and large-scale estimates that are
particularly useful for global ecology applications. An overview of these methods
also serves to reinforce some of the basic principles of primary production
explained earlier.
The basic principle behind remote-sensing approaches is that indices of vegetation structure, especially leaf area index (LAI), are directly related to the photosynthetic capacity of the earths surface. Passive sensors mounted on satellites, such
as the Moderate Resolution Imaging Spectroradiometer (MODIS) instrument,
detect solar radiation reflected from the earths surface, and the ratio of particular
wavelength bands that are differentially absorbed by foliage is quite closely related
to the LAI, at least below some saturation threshold (about LAI 4). This remotely
sensed fraction of the absorbed PAR (APAR absorbed photosynthetically active
radiation) is then used to estimate maximum GPP, and light-use efficiency (LUE) or
production efficiency models adjust for suboptimal environmental conditions and
varying respiratory costs. In particular, the maximum conversion of APAR into
GPP (i.e., LUE) varies among vegetation types because of differences in the size of
plants and consequent total leaf respiration. The LUE will also be reduced by
environmental factors that cause stomatal closure, especially subfreezing temperatures, dry soil, and low atmospheric humidity. The production efficiency will
further depend on the respiratory costs of growing and maintaining all the other
plant tissues.
216
a 800
700
ANPP (g m-2)
217
600
400
600
500
400
200
0
A
300
200
100
0
700
400
600
200
500
400
300
200
100
0
ANPP CV (%)
70
30
60
20
50
10
40
0
O
30
20
10
Se
i-a
rid
G
Ar ras
ct sl
ic an
T d
H und
r
Al ot
pi De a
ne s
e
T r
M H und t
es ot
r
C ic De a
on G s
ife ras er
ro sl t
M us and
M esi For
es c
es
D ic G Old t
ec
f
i
i ra ed
D duo ssl
ec u an
id s F d
uo o
us res
Fo t
re
st
218
219
Fig. 6 Left: View of an annually burned tallgrass prairie watershed at the Konza Prairie Biological Station in Kansas (Photo is taken from a lowland topographic position looking to the uplands).
Right: Spatial variation in aboveground NPP (g/m2) in this watershed. Despite similar plant
communities and minimal climatic variation at this scale (the watershed is about 1 0.5 km in
size), aboveground NPP varies fourfold with highest values in the lowlands near ephemeral
streams (dark blue). Such dramatic variability can be attributed to differences in soil depth,
fertility, and microclimate (From Nippert et al. 2011) (Photo credit: Melinda D. Smith)
10
10
15
20
5
25
30
200
400
600
800
1000
500
1500
0.2
0.0
0.2
0.4
0.6
Precipitation (mm
1000
Sensitivity
ANPP (gm2)
1000
2000
yr1)
2000
3000
2500
Site-based slope
3000
Fig. 7 Global scale relationships between NPP and mean annual temperature (left) and annual precipitation (right). The positive NPPtemperature
relationship is evident only at very large scales (Modified from Schuur 2003). Within biomes, there is often no relationship or the relationship may be
negative since warmer temperatures may lead to greater plant water stress. In contrast, the positive NPPprecipitation relationship is often detected at global,
regional, biome, and individual site scales. In this figure, the curvilinear relationship is from Huxman et al. (2004) based on biomes that ranged from deserts to
tropical forests. The dashed line is the relationship developed by Sala et al. (1988) for grasslands across the central USA. Inset: Huxman et al. (2004)
hypothesized that the sensitivity of ecosystems to changes in precipitation would depend on mean annual precipitation. Ecosystems that receive mean annual
precipitation <1,000 mm are expected to be more sensitive to changes in precipitation amount than ecosystems that have greater mean annual precipitation
15
50
100
150
200
250
300
220
A.K. Knapp et al.
221
Fig. 8 Global relationship between actual evapotranspiration (AET) and NPP (Modified from
Rosenzweig 1968). Because AET combines temperature and precipitation, it is often the single
best predictor of NPP at global scales
conditions throughout its range contributing to dominance by a particular vegetation type. With this temporal approach, there is much more evidence for widespread precipitation than temperature limitation to NPP with the degree of water
limitation varying among biomes. For example, when correlations between year-toyear fluctuations in annual precipitation and variations in aboveground NPP were
compiled for multiple ecosystems across many biomes, the increase in NPP per unit
increase in precipitation in any given year (sensitivity in Fig. 7 inset) was, perhaps
not surprisingly, much greater for drier ecosystems than wetter biomes. In the
wettest of biomes, too much rainfall can become limiting to NPP through either
water-logged soils leading to stressful environments for roots of higher plants (due
to oxygen limitations or anoxia) or extended periods of cloud cover leading to light
limitations to NPP. The most extreme example is in tropical cloud forests where this
combination of constraints leads to a syndrome of stunting of tree size and low
aboveground productivity (Fig. 9).
When a spatial approach to assessing abiotic controls is taken within a biome,
soil fertility and soil depth often referred to as edaphic factors are most often
identified as controlling spatial variation in NPP within many grasslands or forests.
Among a multitude of potential edaphic factors, soil nitrogen availability most
commonly limits terrestrial production with the degree of limitation inferred from
experiments where N is added. Analyses of such experiments in multiple biomes
suggest that additional N alone increases NPP in all but desert biomes (Fig. 10), the
latter being so severely water limited that additional N has little effect. But even in
deserts, there is evidence that in relatively wet years, N addition can increase NPP.
222
TROPICAL MONTANE
ENVIRONMENT:
TMCF Syndrome:
Cool &
Wet
Slow decomposition
& nutrient recycling
Nutrient
limitation
Cloudy
Low leaf
area
index
Saturated
Soil
Windy
Structural
& stability
problems
Shallow
Soil
Frequent
Fog
Low
Aboveground
Net Primary
Productivity
Epiphytic
bryophyte
Loading
High
root:shoot
ratio
Tree stunting,
High Diameter:
Height ratio
Fig. 9 Top. A conceptual model describing the mechanisms whereby several key environmental
factors combined lead to low aboveground productivity and forest stature in tropical montane
cloud forests. TMCF tropical montane cloud forest. Bottom: Photo within a cloud forest in the
Dominican Republic. Note the tree trunks covered with epiphytes (plants that grow on other
plants but are not parasites). Most of the epiphytes are mosses (bryophytes) (Photo credit: Ruth
Sherman)
223
Response Ratio
(ANPP fertilized/ANPP control)
60
50
40
0
0
30
100
300
200
400
500
-1
Mean annual precipitation (m m yr )
20
NS
10
es
er
nd
la
es
fo
r
e
ra
t
Te
m
pe
W
et
t
ca
pi
Tr
o
ca
lg
ra
s
sl
lf
or
an
es
dr
a
an
sl
Tu
n
pi
Tr
o
Te
m
pe
ra
t
gr
as
ve
ra
l
224
Fig. 11 (Top) View of arctic tundra at the Toolik Lake Field Station with experimental infrastructure to assess effects of warming (white structures) and light availability (black structures) on
tundra plant communities. Warming air temperatures alone by 4 C had little effect on ANPP, but
warming soil temperatures, which increased soil N or simply adding N fertilizer increased ANPP
dramatically. (Bottom left) Example of plant communities outside greenhouse structures. (Bottom
right) Increase in biomass and ANPP due to shrub growth inside greenhouse (Photo credit: Alan
K. Knapp)
processes and severely reduced microbial activity that maintains nutrients in forms
unusable for plants. Experiments have been conducted that have warmed arctic
tundra plants or boreal forest trees without warming the soil, and these have
demonstrated that there is little impact on plant growth. But if the soil is warmed,
or if nutrients are added directly, then much greater NPP responses are measured
and in the case of arctic tundra, highly productive shrubs may increase and replace
the previously dominant herbaceous vegetation (Fig. 11). Moreover, warmer summer temperatures may decrease growth in some boreal forests as a result of
increased plant water stress. This interaction between warmer temperatures and
water availability is one that will be discussed later when future controls on NPP in
a world of climate change are considered.
Finally, another useful approach for identifying the climatic factors that are most
likely to limit NPP is to assess limitation of a single vegetation type such as forests
along transects that capture gradients in several potentially limiting factors. One
such study investigated climatic constraints on NPP in coniferous forests along a
transect from coastal Oregon over the Cascade Mountains to western Oregon
(Runyon et al. 1994). Across this transect an eightfold range of NPP was observed,
225
226
respiratory costs of growing and maintaining all the plant tissues that support the
photosynthetic activity in the canopy. In sites with low soil resource availability, a
higher proportion of net photosynthesis must be allocated to the roots to compete
for and acquire water and mineral nutrients. As a result the relative respiratory costs
are much higher on dry and infertile sites. For example, across the Oregon transect,
the fraction of NPP allocated belowground ranges from about 20 % to over 60 %;
high proportional allocation further constrains NPP by increasing the respiratory
costs of the vegetation.
227
3000
Precipitation CV
20
1500
Regional
Continental
Global
400
800
ANPP (g/m2)
Precipitation CV
ANPP
Range of
maximum
ANPP
variability
1200
1600
0
2000
Fig. 13 Relationship between mean annual precipitation (MAP) and ANPP (solid line) and the
coefficient of variation (CV) of precipitation and MAP (dashed line). The precipitation CV is an
indication of how high year-to-year variability is for precipitation. Arid ecosystems generally have
the highest CV and ecosystems with high MAP experience much less interannual variability.
Although ANPP is strongly related to precipitation amount, interannual variability in ANPP is not
highest where the CV of precipitation is highest. Instead, three independent analyses have
concluded that the greatest variability in ANPP from year-to-year peaks between 400 and
800 mm MAP (stacked rectangles). Global analysis by Jung et al. 2011 (maximum ANPP
variability 2501,000 mm), continental by Knapp and Smith 2001 (350850 mm), regional
by Paruelo et al. 1999 (450700 mm)
228
Fig. 14 Evidence for climatic legacy effects of the past years precipitation amounts on the
current years ANPP. Left panel: Two relationships between current years precipitation and
ANPP in semiarid shortgrass steppe grasslands of Colorado. These differ depending on the
previous years ANPP. Note that if the previous year had high levels of ANPP likely due to
high precipitation then the current year has higher ANPP than occurs when the previous year had
low ANPP and precipitation. These relationships were derived from long-term data on interannual
variability in ANPP and precipitation (Redrawn from Oesterheld et al. 2001). Right panel: Results
from an experiment in the Patagonia grassland of Argentina where rainfall inputs into plots were
reduced by 80 %, and then the next year, these same plots provided average and above-average
precipitation amounts and measured ANPP ( filled circles). The solid line represents the general
relationship between ANPP and precipitation developed for this grassland. Note that ANPP was
significantly reduced the year following experimentally imposed drought compared to this general
relationship (Modified from Yahdjian and Sala 2006)
plants including their ability to store carbon for growth in future years, as well as
increases or decreases in stored soil moisture that may extend beyond the current
year. For example, a series of wetter than average years in shrublands of the
Chihuahuan Desert of southern New Mexico led to higher ANPP than would be
expected based on current years precipitation. The mechanisms proposed were that
a series of wet years allowed for an increase in grass cover and abundance within
shrublands, increasing productivity for a given level of rainfall (Peters et al. 2012). A
series of dry years has the opposite effect with a decrease in grasses resulting and
productivity declining. Results from short-term experiments have also identified the
role of meristem limitation (Knapp and Smith 2001) in determining the magnitude
of legacy effects on ANPP. Yahdjian and Sala (2006) experimentally reduced rainfall
inputs by as much as 80 % into a semiarid Patagonian grassland and then measured
ANPP in these previously droughted plots and compared these values to adjacent (
control) plots that had not experienced drought (Fig. 14). They found that ANPP was
reduced by 2030 % due to this drought legacy which reduced the density of plants in
the ecosystem. Thus, when rainfall was returned to normal or even above-average
levels in these previously droughted plots, the density of meristems (growing points
in plants) was much lower than in control plots and this limited NPP.
Pre-hurricane
Hurricane
One year post-hurricane
5
2
er
ov
em
be
r
D
ec
em
be
r
ob
0
ct
Disturbance event
LAI
m
be
0
0
st
Se
p
Au
gu
LAI
LAI
229
te
Month
Fig.15 Left. Relationship between stand age of a temperate forest in Michigan and leaf area index
(LAI, m2 leaf area/m2 ground area) (Modified from Hardiman et al. 2013). Inset: Positive
relationship between LAI and ANPP in forests (Modified from Gower et al. (2001)). Right.
Response of LAI to a hurricane in a scruboak forest. Three years are shown, the year before,
the year of the hurricane, and the recovery year. The hurricane essentially defoliated the stand, but
did not significantly damage twigs, branches, and boles of the trees. Thus, recovery of LAI was
relatively rapid (Modified from Li et al. 2007)
230
grasses as they emerge from the soil in the spring and reduce soil temperature,
slowing rates of nutrient cycling. This large detritus layer can also directly tie up
nutrients making them unavailable to the grasses. The net result is that ANPP is
often much lower in sites where fire has been suppressed for several years compared
to sites in which fire has recently occurred.
Following large-scale disturbances in forest ecosystems, the development of
even-aged stands is marked by a recurring pattern of age-related change in aboveground production. For virtually all cases that have been documented, an early peak
in forest ANPP is followed by a decline as forests get older (Ryan et al. 1997).
Research to explain this phenomenon has been stimulated by its obvious importance
to global forest carbon balance. Despite this apparently universal pattern, a parsimonious explanation does not appear to apply different mechanisms contribute in
various forest types. For example, in some cases, a shift in forest composition from
faster to slower growing species contributes to the temporal pattern; however, the
age-related decline is also observed in forests in the absence of any compositional
change (e.g., monospecific plantations; Fig. 16). Declining production in later stages
of stand development also has been associated with hydraulic constraints on water
transport to the top of tall trees and consequent water stress as well as high respiratory
demands in larger trees. Complex effects of canopy architecture and efficient
utilization of light resources also have been cited as contributing to the age-related
production decline. The implications for optimizing the provision of forest ecosystem services in a changing world continue to be explored.
231
Fig. 17 Contrasting models of the relationship between ANPP and mean annual precipitation
(MAP) based on large-scale spatial data sets (solid line) versus multiyear data from single sites
(dashed lines). Note that in most cases, the slope of the relationship or sensitivity of changes in
ANPP with changes in MAP is less when relationships are based on data from a single site
through time
productivity. Biotic controls on NPP can take many forms besides disturbance or
the previous year(s) being wet or dry. Below four examples of how biotic control on
NPP has been of interest to ecologists for many years have been highlighted.
232
those species with high potential ANPP are not present. For example, in wet years
in semiarid grasslands, where the plant community is dominated by a low density of
short-statured bunch grasses, ANPP will never be as high as in a more mesic
grassland with a high density of very productive grasses even at the same level
of rainfall. This has been termed a vegetation structure constraint on productivity
(Lauenroth and Sala 1992). Interestingly, the temporal relationship between ANPP
and MAP for an individual site predicts higher ANPP in dry years than does the
spatial relationship. This also could be related to differences in vegetation structure
influencing ANPP as well as carry-over of soil moisture from previous years (see
Legacy Effect above). This effect would be captured within relationships at the
site level but not within spatial relationships.
233
1000
Germany
1000
800
600
60
CV of ANPP
ANPP (gm2)
UK
UK
Portugal
400
Sweden
55
50
45
600
Medium soil fertility
400
40
35
200
2
ANPP (gm )
Switzerland
Ireland
800
Greece
5 10 15 20 25
Number of Species
200
0
0
10
20
40
30
Number of Species
10
Number of Species
400
350
300
2
ANPP (gm )
250
1.0
200
Relative frequency
ANPP (gm2)
300
100
0.8
0.6
0.4
150
100
0.2
0.0
10
20
30 40
Species rank
50
50
0
6
200
10
12
14
Number of Species
16
18
10
12
14
16
18
20
Number of Species
Fig. 18 Top: (Left) Evidence supporting the idea that increasing biodiversity (number of species)
of plant communities results in an increase in ANPP and (inset) reduced year-to-year variability
in ANPP (Modified from Loreau et al. 2001). (Right) How soil fertility influences the biodiversityANPP relationship note that when soil nutrients are very low, increasing species
richness has very little effect on ANPP (Modified from Fridley 2002). In both top figures,
these relationships were derived from experimental grassland plant communities assembled
with different numbers of species from a larger pool of potential species. Bottom: (Left) Evidence
for no relationship between ANPP and the number of species in plots in a natural grassland
community. Open and filled circles denote data from two different years of the experiment.
In this study, the number of species was varied by removing species from intact native
grassland plots. Species were selected for removal based upon their abundance with the least
abundant species removed first. Thus, in the inset depicting species ranked by their abundance
(relative frequency of occurrence in plots ~0.81.0 for the most abundant or dominant species),
only those less common to rare species (subordinate species closed symbols) were removed.
(Right)Contrasting responses of the ANPP of dominant (circles) versus subordinate species
(triangles) to number of species (open and closed symbols denote data from two different
years). Note that ANPP of the dominants increased as species number decreased. But production
of the subordinate species increased with greater numbers of species (Modified from Smith and
Knapp 2003)
234
will have more stable ANPP levels from year to year as precipitation varies
compared to plots with only one these other species. When multiple plant species
are present, some also may facilitate the growth of others. Taller species may
provide shade and protection from extreme environmental conditions for shorter
species in a desert environment, for example. These and other mechanisms have
been proposed as responsible for the increase in ANPP with increased plant species
richness. Although this richnessANPP relationship is relatively robust and has
been demonstrated in many parts of the world, it is likely to be most important in
areas with abundant resources. If resources are low, opportunities for sharing and
subdividing resources among many species will also be rare and this effect of
resource availability on the richnessANPP relationship has been experimentally
demonstrated (Fig. 18 top).
The approach taken by ecologists in most experimental tests of the relationship
between richness and ANPP is to construct what has been termed synthetic
communities. These are assembled from random combinations of 1, 2, 4, 8,
16, etc. species and replicated many times in a uniform garden environment.
This approach has strong statistical rationale, but unfortunately most of the
resulting communities are quite dissimilar from natural communities. Natural
communities are not random assemblages of the species that can exist in an area.
Instead, they are more likely to have a few dominant species that make up a large
portion of biomass in every plot. Indeed, if one samples hundreds of plots in a
natural community, dominant species are typically found in all plots sampled. Thus,
synthetic communities, in which each species has an equal chance to be present in
each plot, do not reflect the structure of natural communities where some species
are very common and many are less common or rare (Fig. 18 bottom). Another
experimental approach to assessing the richnessANPP relationship is to use
natural communities and vary richness by removing species. In this case, species
are selected nonrandomly with the least abundant species removed preferentially
compared to the more abundant dominant species. This has been termed realistic
species loss because it has been argued that uncommon species would have the
greatest chance of disappearing from communities. In these studies, no relationship
between richness and ANPP is detected (Fig. 18 bottom), and productivity of the
dominant species, which are present in all plots in these experiments, may actually
increase with reduction in overall community richness. Only within the subordinate
species (those that are not dominant, Fig. 18 bottom) has a positive richnessANPP
relationship been observed (Fig. 18 bottom). The latter suggests that complementarity may be important within this group of species, but because the dominant
species produce so much more biomass than these subordinate species, their
response to richness has little impact on overall ANPP.
Thus, research suggests that species loss can reduce ANPP, particularly if a
dominant species is lost or if species numbers become very low. Furthermore,
mechanisms such as complementarity do operate to increase ANPP as species
numbers increase, but the magnitude of this effect may be small.
1400
235
1200
Young Juniperus stand3
1000
800
Mature Juniperus stand3
600
400
200
0
Grassland1
Initial Shrub Invasion1
5 yrs
20 yrs
35 yrs
80 yrs
Fig. 19 How ANPP in a central US grassland changes when shrubs and then trees replace the
grassland. In this grassland with relatively high levels of precipitation (>800 mm/year), frequent
fire is necessary to maintain grass dominance. If fire is suppressed in this ecosystem (time 0),
grassland ANPP (dominated by Andropogon gerardii) initially decreases, and shrubs (typically
Cornus drummondii or dogwood) that are typically present only as isolated and small individuals
increase dramatically in abundance and cover. These eventually form dense shrub islands that
shade the grasses and eliminate them. This shrub island stage can be the very productive (threefold
higher than the grassland), but eventually even taller woody plants (Juniperus virginiana or
eastern red cedar) displace the shrubs and a forest develops with ANPP ~50 % higher than the
original grassland (Data are from Heisler et al. (2004), Lett et al. (2004), and Norris et al. (2001))
236
Fig. 20 The relationship between grazing intensity and NPP expressed relative to NPP in the
absence of grazing. Shown are four potential relationships. Grazing at any intensity may decrease
NPP indicating that the plant community is unable to replace the tissue lost (undercompensation).
Communities may be able replace tissue lost at low levels of grazing (compensation) but not at
high intensities. Or compensation may occur at all grazing intensities. Finally, there may be low to
moderate levels of grazing where NPP is higher than in areas not grazed (overcompensation)
(Modified from Hilbert et al. 1981)
237
238
Table 2 Belowground primary production in four terrestrial biomes and its relation with total net
primary production. Units are g/m2/year (Adapted from Tierney and Fahey 2007)
Biome
Grassland
Boreal evergreen forest
Temperate deciduous forest
Temperate evergreen forest
Mean BNPP
498
312
380
426
BNPP:NPP
0.52
0.4
0.26
0.24
Table 3 Soil respiration, aboveground litter fall and belowground carbon allocation in three
global forest biomes. Units are g/m2/year (Adapted from Davidson et al. 2002)
Forest biome
Tropical evergreen
Temperate
deciduous
Temperate
evergreen
Soil
respiration
1,603
840
809
Belowground carbon
Litterfall allocation
410
1,193
186
654
Litterfall:
BCA
0.34
0.28
188
0.3
621
(TC, year1) of fine roots in grasslands is higher than in forests so that more rapid
replacement of fine roots contributes significantly to the higher BNPP. The causes
of the higher TC in grasslands are not known, but TC also seems to increase in
warmer and more productive grassland environments. In general, the lifespan of
fine roots (inverse of TC) probably decreases as a result of higher metabolic
activity, nutrient uptake rates, and herbivory in warmer, more fertile soils, thereby
contributing to higher fine root production.
As noted earlier, accurate measurement of BNPP is difficult; however, insights
into the process have been provided by the straightforward measurement of TRA as
the difference between total soil respiration (TSR) and aboveground litterfall flux of
C. On average in the worlds forests, annual TSR is about three times greater than
aboveground litterfall, and hence, TRA is about twice as large as aboveground
litterfall. A synthesis for temperate forests also indicates that TRA is about three
times greater than BNPP, implying that about two-thirds of TRA is used in root
respiration (Table 3). Measurements of TRA also provide a basis for evaluating
BNPP responses to varying soil resource availability. In many temperate zone
ecosystems, nitrogen is the most growth-limiting soil nutrient, and the availability
of N has increased markedly as a result of anthropogenic activity, with likely
consequences for NPP. Increasing soil N availability might be expected to cause
a decrease in fine root biomass, but at the same time, it could stimulate higher fine
root turnover. The functional equilibrium hypothesis would argue for a reduction in
proportional TRA in more fertile soils, and some evidence supports this conjecture
(Fig. 21). A further complication is that a large proportion of BNPP goes to RCF.
For example, Jones et al. (2009) summarized available evidence to estimate that an
average of 27 % of TRA goes to RCF, equivalent to 11 % of net photosynthesis.
These observations emphasize the importance of RCF in facilitating soil resource
239
800
700
600
500
400
0
Fig. 21 Relationship between belowground carbon allocation and soil nitrogen availability
(as indicated by net nitrification) in six northern hardwood forests in New Hampshire. Note that
as soil fertility increases, estimated carbon allocated belowground decreases (Kikang Bae,
unpublished data)
240
241
Fig. 23 (Left) Interaction between mean annual temperature (MAT) and response of ANPP to
experimental warming. This was derived by combining results from 127 different warming
experiments (warming magnitude ranged from 1 C to 5 C) across sites that varied widely in
MAT (Modified from Lin et al. 2010). (Center) Response of NPP to warming under conditions of
either high or low precipitation modeled for 7 ecosystems ranging from annual grassland to
tropical and boreal forests (Modified from Luo et al. 2008). (Right) Increase in ANPP with
increasing MAT for North America coastal marshlands dominated by Spartina alterniflora.
Note the strong effects of increased temperature in an ecosystem that always has abundant water
(Modified from Kirwan et al. 2009)
For example, at high temperatures, respiration may increase more than photosynthesis to warming leading to reduced NPP. In addition, unless precipitation inputs
are very high, ecosystems with high MAT are likely to experience substantial water
stress (due to high evapotranspiration) and considerable evidence indicates that
warming effects on NPP can be strongly influenced by water availability. In studies
in which both precipitation and warming are varied, warming effects are positive
with high precipitation but negative with low precipitation (Fig. 23). The latter
response is interpreted as evidence that increased water stress in plants caused by
warming has a much stronger negative effect than any positive effects of warming.
The importance of this interaction between temperature and water availability can
be further demonstrated by assessing temperature effects on NPP in wetland
ecosystems where water is never limiting. Recall that for most biomes, interannual
variability in temperature usually does not correlate with year-to-year variation in
NPP. This is in sharp contrast to precipitation where wet vs. dry years results in high
and low ANPP, respectively (see Abiotic Controls on NPP section above). The
exception to this pattern is for wetlands where there can be a strong temperature
response by ANPP over time (warmer years have higher ANPP) and space (wetlands with higher MAT at lower latitudes are more productive, Fig. 23).
The modifying effect of water availability on NPP responses to warming is also
evident when assessing NPP responses to increased CO2 in the atmosphere. As
noted earlier, evidence is overwhelming that many climatic changes (including
warming) can be attributed to the 25 % increase in atmospheric CO2 measured in
the last 50 years with even greater climate change forecast for the next 100 years.
Because CO2 is, of course, essential for photosynthesis and NPP, ecologists have a
long history of assessing the impacts of increased CO2 on key ecological processes
such as NPP. In general, experiments that have increased CO2 to individual plants,
242
Fig. 24 Examples of how the effect of increased levels of atmospheric CO2 on ANPP depends on
water availability. (Left) Relationship between the increase in ANPP due to increased CO2
(calculated as the proportional increase in NPP in ecosystems with ~600700 ppm CO2 relative
to ambient levels) and growing season rainfall. This relationship was developed from the results of
independent field experiments in grasslands in Colorado, Kansas, and Switzerland. Note that in
grasslands with low rainfall during the growing season, elevated CO2 increases ANPP by up to
50 %. But this enhancement is small in grasslands that are very wet during the summer. (Middle)
Even in a single grassland with relatively high rainfall (Switzerland), the amount of rain that fell
the preceding 6 weeks prior to harvest of biomass can have a strong impact on how much elevated
CO2 increases ANPP. (Right) Semiarid grasslands in Wyoming and Colorado show an even
greater sensitivity of CO2 responses to soil moisture levels (Left and middle panels modified
from Morgan et al.2004; Right panel modified from Morgan et al. 2011)
and entire ecosystems, report that elevated CO2 increases NPP. This can be due to
the direct stimulation of photosynthesis by the greater availability of CO2
(a CO2fertilization mechanism) as well as because stomatal opening in plants is
almost always reduced when CO2 is increased; this reduces transpiration and
improves plant water status which can also increase photosynthesis, growth, and
NPP (a water conservation mechanism). Field studies in grasslands clearly
demonstrate the interaction between water and the effect of increased CO2
(Fig. 24). Grasslands dominated by C4 species show this interaction with water
most strongly (Wyoming and Colorado grasslands in Fig. 24) since the C4 pathway
is generally not subject to CO2 fertilization (CO2 concentrations are very high in the
bundle sheath cells inside leaves) but stomatal sensitivity to CO2 is still evident. In
this case, CO2 stimulation of ANPP only occurs under dry conditions (when water
conservation matters) and not when moisture is plentiful. However, even in
C3-dominated grasslands (Swiss grasslands in Fig. 24) where direct CO2 fertilization can increase NPP, the positive effects on NPP of water conservation at high
CO2 also are evident.
The high costs of conducting global change experiments, particularly those that
alter CO2, temperature, or precipitation, have resulted in a preponderance of singlefactor studies. This is unfortunate because NPP in the future will be determined by
multiple global change drivers impacting ecosystems concurrently. Experiments
that alter single factors can certainly provide mechanistic insight for how NPP
might respond to a change in a global change driver (see Figs. 10, 14, 18, 23, 24),
and they can be quite valuable for parameterizing simulation models. They can also
243
Fig. 25 Response of NPP of five sweetgum (Liquidambar styraciflua) forest stands in Tennessee
grown under two CO2 levels. Open circles display mean values of stands grown under ambient
CO2 levels (~380 ppm), while the dark circles show mean biomass production under elevated
levels of CO2 (~550 ppm). Note that the stimulation of NPP that occurred initially diminished over
time. Despite projections of higher CO2 levels in the next century due to climate change, forest
NPP may be constrained by other limiting factors, such as nitrogen availability (Modified from
Norby et al. 2010)
244
Future Directions
Better quantification of the proportion of total NPP that goes to belowground
NPP, in particular rhizosphere carbon flux including exudation, rhizodeposition,
and allocation to mycorrhizal fungi and other symbionts.
Determination of how increasing human activities will influence disturbance
regimes, land-use patterns, and vegetation structure, pattern, and composition
(e.g., through the introduction of exotic species) and consequently affect NPP.
Increased understanding is needed regarding how global environmental changes
such as warming temperatures, changing precipitation regimes, increased atmospheric CO2, and greater rates of nitrogen deposition will influence global NPP
in the future. Much is known of how many of these will affect NPP in individual
ecosystems from single-factor experiments, but their combined effects across
multiple ecosystems are highly uncertain.
References
Davidson EA, Savage K, Bolstad P, Clark DA, Curtis PS, Ellsworth DS, Hanson PJ, Law BE,
Luo Y, Pregitzer KS, Randolph JC, Zak D. Belowground carbon allocation in forests estimated
from litterfall and IRGA-based soil respiration measurements. Agr Forest Meteorol.
2002;113:3951.
Fridley JD. Resource availability dominates and alters the relationship between species
diversity and ecosystem productivity in experimental plant communities. Oecologia.
2002;132:2717.
Gower ST, Krankina O, Olson RJ, Apps M, Linder S, Wang C. Net primary production and carbon
allocation patterns of boreal forest ecosystems. Ecol Appl. 2001;11:1395411.
Haberl H, Erb KH, Krausmann F, Gaube V, Bondeau A, Plutzar C, Gingrich S, Lucht W, FischerKowalski M. Quantifying and mapping the human appropriation of net primary production in
earths terrestrial ecosystems. Proc Natl Acad Sci USA. 2007;104:129427.
Hardiman BS, Gough CM, Halperin A, Hofmeister KL, Nave LE, Bohrer G, Curtis
PS. Maintaining high rates of carbon storage in old forests: a mechanism linking canopy
structure to forest function. For Ecol Manage. 2013;298:1119.
Harpole WS, Ngai JT, Cleland EE, Seabloom EW, Borer ET, Bracken MES, Elser JJ, Gruner DS,
Hillebrand H, Shurin JB, Smith JE. Nutrient co-limitation of primary producer communities.
Ecol Lett. 2011;14:85262.
Heisler JL, Briggs JM, Knapp AK, Blair JM, Seery A. Direct and indirect effects of fire
on shrub density and aboveground productivity in a mesic grassland. Ecology.
2004;85:224557.
Hilbert DW, Swift DM, Detling JK, Dyer MI. Relative growth rates and the grazing optimization
hypothesis. Oecologia. 1981;51:148.
Huston MA, Wolverton S. The global distribution of net primary production: resolving the
paradox. Ecol Monogr. 2009;79:34377.
Huxman TE, Smith MD, Fay PA, Knapp AK, Shaw MR, Loik ME, Smith SD, Tissue DT, Zak JC,
Weltzin JF, Pockman WT, Sala OE, Haddad BM, Harte J, Koch GW, Schwinning S, Small EE,
Williams DG. Convergence across biomes to a common rain-use efficiency. Nature.
2004;429:6514.
Jones DL, Nguyen C, Finlay RD. Carbon flow in the rhizosphere: carbon trading at the soilroot
interface. Plant and Soil. 2009;321:533.
245
Jung M, Reichstein M, Margolis HA, Cescatti A, Richardson AD, Arain MA, Arneth A,
Bernhofer C, Bonal D, Chen J, Gianelle D, Gobron N, Kiely G, Kutsch W, Lasslop G, Law
BE, Lindroth A, Merbold L, Montagnani L, Moors EJ, Papale D, Sottocornola M, Vaccari F,
Williams C. Global patterns of land-atmosphere fluxes of carbon dioxide, latent heat, and
sensible heat derived from eddy covariance, satellite, and meteorological observations. J
Geophys Res. 2011;116:G00J07.
Keeling HC, Philips OL. The global relationship between forest productivity and biomass. Glob
Ecol Biogeogr. 2007;16:61831.
Kirwan ML, Guntenspergen GR, Morris JT. Latitudinal trends in Spartina alterniflora
productivity and the response of coastal marshes to global change. Glob Chang Biol.
2009;15:19829.
Knapp AK, Smith MD. Variation among biomes in temporal dynamics of aboveground primary
production. Science. 2001;291:4814.
Knapp AK, Briggs JM, Childers DL, Sala OE. Estimating aboveground net primary production in
grassland and herbaceous dominated ecosystems. In: Fahey TJ, Knapp AK, editors. Principles
and standards for measuring net primary production. New York: Oxford University Press;
2007. p. 2748.
Lauenroth WK, Sala OE. Long-term forage production of North American shortgrass steppe. Ecol
Appl. 1992;2:397403.
LeBauer DS, Treseder KK. Nitrogen limitation of net primary productivity in terrestrial ecosystems is globally distributed. Ecology. 2008;89:3719.
Lett MS, Knapp AK, Briggs JM, Blair JM. Influence of shrub encroachment on plant productivity
and carbon and nitrogen pools in a mesic grassland. Can J Bot. 2004;82:136370.
Li J, Powell TL, Seiler TJ, Johnson DP, Anderson HP, Bracho R, Hungate BA, Hinkle CR, Drake
BG. Impacts of Hurricane Frances on Florida scrub-oak ecosystem processes: defoliation, net
CO2 exchange and interactions with elevation CO2. Glob Chang Biol. 2007;13:110113.
Lin D, Xia J, Wan S. Climate warming and biomass accumulation of terrestrial plants: a metaanalysis. New Phytol. 2010;188:18798.
Loreau M, Naeem S, Inchausti P, Bengtsson J, Grime JP, Hector A, Hooper DU, Huston MA,
Raffaelli D, Schmid B, Tilman D, Wardle DA. Biodiversity and ecosystem functioning: current
knowledge and future challenges. Science. 2001;294:8048.
Luo Y, Gerten D, Le Maire G, Parton WJ, Weng E, Zhou X, Keough C, Beier C, Ciais P,
Cramer W, Dukes JS, Emmett B, Hanson PJ, Knapp A, Linder S, Nepstad D, Rustad
L. Modeled interactive effects of precipitation, temperature, and CO2 on ecosystem carbon
and water dynamics in different climatic zones. Glob Chang Biol. 2008;14:198699.
Morgan JA, Pataki DE, Korner C, Clark H, Del Grosso SJ, Gr
unzweig JM, Knapp AK, Mosier AR,
Newton PCD, Niklaus PA, Nippert JB, Nowak RS, Parton WJ, Polley HW, Shaw MR. Water
relations in grassland and desert ecosystems exposed to elevated atmospheric CO2. Oecologia.
2004;140:1125.
Morgan JA, LeCain DR, Pendall E, Blumenthal DM, Kimball BA, Carrillo Y, Williams DG,
Heisler-White J, Dijkstra FA, West M. C4 grasses prosper as carbon dioxide eliminates
dessication in warmed semi-arid grassland. Nature. 2011;476:2025.
Nippert JB, Ocheltree TW, Skibbe AM, Kangas LC, Ham JM, Shonkwiler Arnold KB, Brunsell
NA. Linking plant growth responses across topographic gradients in tallgrass prairie.
Oecologia. 2011;166:113142.
Norby RJ, Warren JM, Iversen CM, Medlyn BE, McMurtie RE. CO2 enhancement of forest
productivity constrained by limited nitrogen availability. Proc Natl Acad Sci USA.
2010;107:1936873.
Norris MD, Blair JM, Johnson LC, McKane RB. Assessing changes in biomass, productivity, and
C and N stores following Juniperus virginiana forest expansion into tallgrass prairie. Can J For
Res. 2001;31:19406.
Oesterheld M, Loreti J, Semmartin M, Sala OE. Inter-annual variation in primary production of a
semi-arid grassland related to previous-year production. J Veg Sci. 2001;12:13742.
246
Paruelo JM, Lauenroth WK, Burke IC, Sala OE. Grassland precipitation-use efficiency varies
across a resource gradient. Ecosystems. 1999;2:648.
Peters DPC, Yao J, Sala OE, Anderson JP. Directional climate change and potential reversal of
desertification in arid and semiarid ecosystems. Glob Chang Biol. 2012;18:15163.
Rosenzweig M. Net primary productivity of terrestrial environments: predictions from climatological data. Am Nat. 1968;102:6774.
Runyon J, Waring RH, Goward SN, Welles JM. Environmental limits on net primary production
and light-use efficiency across the Oregon transect. Ecol Appl. 1994;4:22637.
Ryan MG, Binkley D, Fownes JH. Age-related decline in forest productivity: pattern and process.
Adv Ecol Res. 1997;27:21362.
Sala OE, Parton WJ, Joyce LA, Lauenroth WK. Primary production of the central grassland region
of the United States. Ecology. 1988;69:405.
Schuur EAG. Productivity and global climate revisited: the sensitivity of tropical forest growth to
precipitation. Ecology. 2003;84:116570.
Smith MD, Knapp AK. Dominant species maintain ecosystem function with non-random species
loss. Ecol Lett. 2003;6:50917.
Tierney GL, Fahey TJ. Evaluating minirhizotron estimates of fine root longevity and production in
the forest floor of a temperate broadleaf forest. Plant and Soil. 2001;229:16776.
Tierney GL, Fahey TJ. Estimating belowground primary productivity. In: Fahey TJ, Knapp AK,
editors. Principles and standards for measuring net primary production. New York: Oxford
University Press; 2007.
Wu Z, Dijkstra P, Koch GW, Penuelas J, Hungate BA. Responses of terrestrial ecosystems to
temperature and precipitation change: a meta-analysis of experimental manipulation. Glob
Chang Biol. 2011;17:92742.
Yahdjian L, Sala OE. Vegetation structure constrains primary production response to water
availability in the Patagonian steppe. Ecology. 2006;87:95262.
Yahdjian L, Gherardi L, Sala OE. Nitrogen limitation in arid-subhumid ecosystems: a metaanalysis of fertilization studies. J Arid Environ. 2011;75:67580.
Zhao M, Running SW. Drought-induced reduction in global terrestrial net primary production
from 2000 through 2009. Science. 2010;329:9403.
Further Readings
Fahey TJ, Knapp AK, editors. Principles and standards for measuring net primary production.
New York: Oxford University Press; 2007.
Gill RA, Kelly RH, Parton WJ, Day KA, Jackson RB, Morgan JA, Scurlock JMO, Tieszen LL,
Castle JV, Ojima DS, Zhang XS. Using simple environmental variables to estimate belowground productivity in grasslands. Glob Ecol Biogeogr. 2002;11:7986.
McNaughton SJ. Compensatory plant growth as a response to herbivory. Oikos. 1983;40:32936.
Norby RJ, Hanson PJ, ONeill EG, Tschaplinski TJ, Weltzin JF, Hansen RA, Cheng W,
Wullschleger SD, Gunderson CA, Edwards NT, Johnson DW. Net primary productivity of a
CO2-enriched deciduous forest and the implications for carbon storage. Ecol Appl.
2002;12:12616.
Running SW, Nemani RR, Heinsch FA, Zhao M, Reeves M, Hashimoto H. A continuous satellitederived measure of global terrestrial primary production. Bioscience. 2004;54:54760.
Smith MD, Knapp AK, Collins SL. A framework for assessing ecosystem dynamics in response to
chronic resource alterations induced by global change. Ecology. 2009;90:327989.
Contents
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Biogeography and Climate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Vegetation Structure and Phenology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Tropical Rain Forest Biodiversity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Why Are There so Many Tree Species in Tropical Forests? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Case Study: Plant Pests Maintain Tree Species Diversity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Productivity and Nutrient Cycling in Tropical Rain Forests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Threats to Tropical Rain Forests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Case Study: Oil Palm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Future Directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
249
250
252
257
259
262
263
265
267
267
269
Abstract
Occupying less than 7 % of Earths land surface, tropical rain forests harbor
perhaps half of the species on Earth and are ecologically, economically, and
culturally crucial for issues in global food security, climate change, biodiversity, and human health.
Geographically located between the latitudes 10 N and 10 S of the equator,
lowland tropical rain forest ecosystems share similar physical structure but
vary in geology, species composition, and anthropogenic threats across the
forests of Southeast Asia, Australia, Africa, and Central and South America.
Mature tropical rain forests are stratified by multiple canopy and understory
layers, and physiognomic properties include evergreen broadleaf tree species,
a preponderance of species with large leaves to aid with sunlight capture in
the light-limited understory, and leaf properties such as entire margins and
drip tips that channel water efficiently from the leaf surface.
R.E. Gallery (*)
School of Natural Resources and the Environment, University of Arizona, Tucson, AZ, USA
e-mail: rgallery@email.arizona.edu
# Springer Science+Business Media New York 2014
R.K. Monson (ed.), Ecology and the Environment, The Plant Sciences 8,
DOI 10.1007/978-1-4614-7501-9_4
247
248
R.E. Gallery
249
Greater use of remote sensing imagery from satellites, airborne Light Detection and Ranging (LiDAR) data, and unmanned drones will allow accurate
tracking of disturbance and C stocks as well as monitoring of phenology,
foliar canopy chemistry, individual species identification, and biodiversity
estimates from local to regional scales.
The tropical biome is undergoing significant change. Understanding
the drivers and impacts of these changes will require sustained advances
across multiple disciplines. Ultimately as a society, we are left asking
what is the capacity of our remaining and regrowing tropical rain
forests to adapt to long-term anthropogenic and climate change and what
can we do to moderate these effects while nourishing a healthy human
population?
Introduction
Along with their extraordinary biodiversity and predominant influences on global
carbon (C), nitrogen (N), and water cycles, tropical rain forests provide
powerful inspiration that has driven biological inquiry for centuries. Theories in
biogeography, ecology, and evolution by natural selection crystallized through
the South America and Southeast Asian journeys of Alexander von Humboldt,
Charles Darwin, Alfred Russel Wallace, and Johannes Eugenius Bulow
Warming considered by some to be the founder of tropical ecology. From the
lowland rain forests of Venezuela into the Andes, von Humboldt recorded the
change in vegetation with climate, drawing the first conclusions that laid the
groundwork for the field of biogeography. Both Darwin and Wallace developed
their ideas of evolution by natural selection through their observations of exceptional species diversity in South America and Southeast Asian rain forests.
Current research questions in tropical rain forest plant ecology comprise determining the origins and maintenance of such extraordinary genetic, species, and
habitat diversity; the factors that regulate net primary productivity (NPP) of intact
and disturbed tropical forests; and the consequences of the loss and conversion of
these forests on global biogeochemical cycles, water cycles, and ecosystem
services.
Occupying less than 7 % of Earths land surface, tropical rain forests harbor
perhaps half of the species on Earth and are ecologically, economically, and
culturally crucial for issues in global food security, climate change, biodiversity,
and human health. Tropical rain forests share a particular combination of climate
parameters, floristic composition, forest structure, and plant physiognomy. Though
they differ in geology and climate patterns such as intensity of El Nino
Southern Oscillation (ENSO) events, tropical rain forests face the common threats
of deforestation, land use conversion, invasive species, and changing climate that
require the same dedication to conservation and management practices that best
suit the unique socioeconomic and cultural characteristics of each region.
250
R.E. Gallery
Current global, multi-institutional networks, such as the Center for Tropical Forest
Science (ctfs.si.edu), monitor the growth and survival of approximately 4.5 million
trees and 8,500 species in forests around the world to understand forest function,
diversity, and sustainable management to inform natural resource policy and build
capacity in the face of climate and land use change.
Fig. 1 Tropical forest distributions and carbon stored by biome (in gigatonnes). Tropical and subtropical forests store more C than any other biome (Reprinted
with permission (Riccardo Pravettoni, UNEP/GRID-Arendal, http://www.grida.no/graphicslib/detail/carbon-stored-by-biome_9082))
9
Ecology of Tropical Rain Forests
251
252
R.E. Gallery
tropical montane forests, and tropical deciduous forests, the environmental variables driving ecosystem processes and plant adaptations such as fog, in the case of
montane forests, and fires and drought in seasonally dry tropical forest are sufficiently different from tropical rain forests and are beyond the scope of this chapter.
Fig. 2 (a) Aerial photo of a Neotropical rain forest canopy. The brilliant yellow crowns display
the synchronous flowering of Tabebuia guayacan (Bignoniaceae) trees. Emergent trees rise above
the forest canopy and palm trees and various tree architectures are apparent. The range in hue of
individual crowns depicts variation in foliar chemistry and water content (Photo credit Christian
Ziegler). (b) Cross section of a lowland Amazon rain forest in Manu National Park, Peru, shows a
distinct vertical profile from understory shrubs to emergent trees. River erosion exposes roots
(Photo credit Kyle Dexter)
253
Fig. 3 (a) The buttress of this Ficus (Moraceae) tree in Corcovado National Park, Costa Rica,
provides support and stability in shallow tropical forest soils (Photo credit Andrea Vincent).
(b) Woody lianas rely on trees for structural support to reach the forest canopy. Liana abundance
and biomass are increasing in a number of tropical rain forests, including the La Selva Biological
Station, Costa Rica, where this photo was taken, with significant implications for tree community
diversity, gap dynamics and forest structure, and tropical forest nutrient cycling (Photo credit
Eloisa Lasso)
leaf properties such as entire margins and drip tips that channel water efficiently
from the leaf surface. Cauliflory, the development of flowers on tree trunks and
main branches, is common in aseasonal tropical understory trees and facilitates
pollination by non-volant insects or animals. Not surprisingly, the percentage of
deciduous tree species increases with increasing seasonality. Across the strong
precipitation gradient along the Isthmus of Panama, deciduous trees account for
less than 5 % in more aseasonal forests on the Atlantic to a quarter of tree species in
the forest communities on the Pacific side.
Competition for light, water, and nutrients varying over heterogeneous landscapes generate and shape ecophysiological adaptations in plants. Equatorial solar
radiation levels are high, and canopy leaves and leaves exposed to direct sunlight
experience very different irradiance and humidity than understory leaves. Greater
than 99 % of sunlight is absorbed and reflected as the light passes through the forest
canopy, resulting in low light intensity and quality in the forest understory where
competition for light is high and certain plants can rapidly respond to the patchworks of light created by sunflecks. Life history strategies across the light demanding to shade tolerant spectrum include, at the one end, pioneer species with high
254
R.E. Gallery
photosynthesis and respiration rates and low wood density to slow growing, welldefended, high wood density species that can persist in the understory until a gap
forms overhead. Species are aligned across a competitioncolonization continuum
along a multitude of axes including seed size and dispersal, leaf lifespan, and
population turnover that together highlight tradeoffs in resource allocation and
reproductive strategies. Water limitation controls transpiration and photosynthesis,
and tropical trees can transpire several hundred liters of water a day, which
emphasizes the importance of reducing cavitation risks during low water availability. Of course environmental tolerances to temperature and water availability drive
global patterns of plant distributions, and within tropical forests interspecific
differences in drought tolerance have been shown to determine plant species
distributions at local scales and across the strong rainfall gradient of the Isthmus
of Panama. Among the soil nutrients that affect plant productivity, phosphorous (P),
which is rapidly mobilized by chemical and microbial activity, is often limiting in
highly weathered tropical soils. A more detailed discussion of biogeochemistry and
plant productivity can be found in section Productivity and Nutrient Cycling in
Tropical Rain Forests.
While lianas are found in temperate rain forests, their predominance and diversity in tropical rain forests are notable, as is the trend that they are increasing in
abundance and biomass in a number of tropical rain forests. Contributing up to
45 % of woody stems and 35 % of species richness in a tropical forest community,
lianas significantly reduce tree growth rates through direct competition, more than
double tree mortality risks, and increase gap size and severity through canopy
connectivity, and the capacity for lianas to alter successional pathways in tropical
rain forests is only beginning to be understood (van der Heijden et al. 2013). An
increase in liana biomass has serious implications for tree community diversity, gap
dynamics and forest structure, and tropical forest nutrient cycling. For example,
lianas reduce tree growth and survival in the slower-growing, higher wood density
trees that support them, which, along with changing gap regimes, shift species
composition towards faster-growing trees with lower wood density. While accurate
predictions require more data, the additive effects of an increase in liana biomass
are correlated with a reduction in tropical forest C storage, a value that is currently
not considered in global vegetation models.
Little is known about cambial phenology the seasonality of stem growth in
tropical rain forest trees. Our lack of understanding of the triggering factors of
cambial dormancy in tropical rain forest trees has lead to the long-standing
assumption that tropical trees do not form annual growth rings (Jacoby 1989;
Worbes 2002). Furthermore, the complex wood anatomy characteristic of the
majority of tropical tree species has long steered dendrochronologists away from
tropical regions. In recent decades, however, distinct annual growth ring boundaries, often consisting of marginal parenchyma bands and induced by cambial
dormancy, have been detected in multiple lowland tropical rain forest species.
As a result, an increasing number of reliable, climate-sensitive tree-ring chronologies are now available based on trees from various tropical biomes across Asia, the
Amazon region, and Africa. These chronologies reflect seasonally fluctuating
255
climatic conditions that typically consist of distinct dry seasons but can also consist
of periodical flooding (Schongart et al. 2004). In regions with a bimodal rainfall
distribution (e.g., eastern Africa), trees can exhibit a bimodal pattern of cambial
activity, and two growth rings can be found per year. Water availability is a major
driver of phenological periodicity in seasonal tropical rain forests, and leaf phenology is generally synchronized with the seasonality of soil water content and tree
water status. In deciduous trees, leaf fall typically occurs at the end of the
dry season and leaf flushing in the wet season. Deciduousness, however, is species
and site specific and can be a function of tree canopy status, with canopy and
emergent trees generally showing a more distinct phenological seasonality and
deciduousness than understory trees (see Fig. 2a). There is plasticity in this trait;
some species have seasonal leaf fall at dry sites but are evergreen at sites with less
moisture stress.
Though the climate of tropical rain forests has more tempered seasonality
relative to other ecosystems, most rain forest tree species do not grow, flower, or
fruit year-round. Periods of leaf flush, bud burst, flowering, fruiting, and senescence
that are related to climate conditions and day length (photoperiod) are considered
phenological responses, the proximate and ultimate causes of which have been
studied from individual variation within populations to community and guild-level
patterns. In seasonal tropical rain forests, peaks in leaf flushing, flowering (see
Fig. 2a), and fruiting coincide with the high irradiance and low water stress
associated with the onset of the wet season. This synchrony of events is common
within communities and largely driven by resource availability, though biotic
explanations for synchrony include selection to attract pollinators or seed dispersers
and to avoid herbivory and seed predation (van Schaik et al. 1993). The synchronous flowering of canopy emergent tree species such as Dipteryx panamensis
(Fabaceae) is visible in high-resolution satellite images, which enable individual
tracking and have revolutionized the study of remote and large tracks of forests.
Synchronous supra-annual flowering and mast fruiting that may lead to seed
predator satiation are defining features of the Dipterocarp forests of Southeast
Asia, with Borneo housing the greatest diversity of Dipterocarpaceae that are
increasingly threatened by extensive logging and land conversion. Bamboos also
wait decades between synchronized flowering before dying back. Monocarpic or
semelparous trees that reproduce only once are uncommon, though examples can be
found in the Neotropical genera Tachigali (Fabaceae) and Spathelia (Rutaceae) and
the genus Harmsiopanax (Araliaceae) in tropical Asia. Wind pollination is relatively rare in tropical rain forests and many coevolutionary pollination, and seed
dispersal relationships have developed between plants and insects, birds, bats, fish,
and mammals.
In this chapter on tropical rain forest plant ecology, I would be remiss not to
highlight a few of the archetypal associations between tropical plants and the
organisms that rely on them for food and habitat. Each of the examples detailed
below are pantropical and emphasize the extraordinary complexity of ecological
systems. They also demonstrate the coevolution of symbiotic relationships between
plants, insects, and fungi for protection, nutrient acquisition, and pollination.
256
R.E. Gallery
257
of any tropical tree species, with dispersal ranges of hundreds of square kilometers
driven by fig wasp-mediated gene flow and seed dispersal via the numerous fig
frugivores.
Mycorrhizal associations between plant roots and symbiotic fungi are pervasive
and not unique to tropical rain forests; greater than 90 % of plant families form
mycorrhizal associations. While ectomycorrhizal tree species are less common,
both endo- and ectomycorrhizal fungi are found in tropical forests worldwide, and
trees can host both groups of symbionts simultaneously. Arbuscular mycorrhizas
(AM; Glomeromycota) are endomycorrhiza whose hyphae enter plant cells and
produce vesicles or arbuscules that increase the surface area of contact between the
plant root and fungus to facilitate nutrient transfer. AM fungi are cosmopolitan with
broad host ranges though different plant species responses to mycorrhiza communities can influence the competitive outcome among seedlings. Ectomycorrhizas
(EM) are found across fungal phyla (Basidiomycota, Ascomycota, Zygomycota)
and their species number in the thousands compared to only hundreds of arbuscular
mycorrhizal species. EM hyphae sheath the root and an extensive hyphal network,
called a Hartig net, runs between plant cells within the root cortex. Tree species
with EM are less common than those with AM, but all species of Dipterocarpaceae
form EM associations, as do species in the Fagaceae and Fabaceae subfamily
Caesalpinioideae. In both types of association, carbon fixed from the plant is
transferred to the heterotrophic fungus. In return both ecto- and endomycorrhizas
increase root surface area, thereby improving plant nutrient acquisition of P, N,
calcium, potassium and other ions that tend to be limiting in tropical soils. There
is evidence that these associations also improve plant resistance to root
pathogens and tolerance to drought. The host-specific effect of different mycorrhizal communities on plant growth has been proposed as a potential mechanism
reducing plant community richness. Tree species hosting particular suites of
mycorrhizal communities could create a positive feedback for conspecific over
heterospecific juvenile recruitment. Furthermore, in certain low diversity forests the
dominant tree species tends to form EM associations and it has been hypothesized
that an EM network may provide recruitment advantages to EM plant species
over non-EM plant species through positive feedbacks. This hypothesis requires
further testing.
258
R.E. Gallery
The current diversity and distribution of modern plant lineages has been shaped by
numerous extinction (e.g., Devonian, Permian, Cretaceous) and radiation events
throughout Earths history. The retraction of tropical rain forests during the cooler,
drier Pleistocene glacial periods (ca. 100,000 year per cycle) and expansion of tropical
rain forests during warmer, wetter interglacial periods (ca. 1020,000 year per cycle)
created fragmented refugia in African and Australian, though recent evidence suggests not Neotropical, forests, that may have promoted lineage differentiation and
allopatric speciation that contribute to the extant high tropical plant diversity. Different scales over which diversity is measured include alpha diversity (local, habitat
scale), beta diversity (species turnover at landscape to regional scales), and gamma
diversity (total regional species richness). Since regional diversity reflects a balance
between speciation and extinction, it should be higher in larger, older areas that offer
more opportunities for isolation and divergence through environmental heterogeneity
as well as lower extinction probabilities through species-area relationships and
millennia without major climatic shifts, in other words, in tropical rain forest biomes.
Continental drift, climate constraints, and long-distance dispersal are responsible for some of the similarities and differences in species across tropical regions.
Dipterocarpaceae are dominant only in Southeast Asia, and palms (Arecaceae) and
legume species in the Fabaceae are abundant in South American tropical rain
forests (e.g., Fig. 2a), but not in African ones. There are, however, a number of
plant families shared between South America, Africa, and Southeast Asia (from
27 to 44 in a recent global analysis of 4 ha plots by Ricklefs and Renner 2012). In
contrast, independent diversification and species radiation mean that much fewer
genera are shared across regions. Between 58 % and 68 % of plant families
(44 families) are shared between Yasuni, Ecuador, (65 families) and Pasoh, Malaysia, (76 families), whereas only approximately 12 % (35 genera) of their
296 (Yasuni) and 259 (Pasoh) genera overlap. Some species are widely distributed
with pantropical ranges, for example, Ceiba pentandra (Malvaceae), a canopy
pioneer tree, whose range encompasses Central and South America, the Caribbean,
and eastern Africa. Interestingly, the low nucleotide divergence in microsatellite
chloroplast and nuclear ribosomal DNA data among Neotropical and African
populations supports long-distance dispersal, and not vicariance, as the explanation
for this species range (Dick et al. 2007). Population genetic data provide a means
of inferring the dispersal and historical biogeography of species. See Kraft and
Ackerly (Chap. 3, Assembly of Plant Communities) for an excellent description of phylogenetic analysis and structure within and among communities.
Tropical rain forests harbor approximately two thirds of the estimated
350,000500,000 extant flowering plant species on Earth. Floristic endemism,
whose cause may be attributed to young species age, is high especially in island
systems such as Indonesia where greater than 50 % of the indigenous vascular plant
taxa do not occur anywhere else. Although tropical rain forests are generally
considered synonymous with diversity, within these systems tree alpha diversity
varies considerably and is broadly correlated with mean annual temperature (MAT)
and mean annual precipitation (MAP). Numerous studies using the CTFS forest
inventory plots reveal that patterns of alpha diversity and species or familial
259
dominance vary across African, American, and Asian tropical rain forests from a
mean of 22 species of tree 10 cm dbh per ha in southern India to 254 species per
ha in Ecuadorian Amazon (Table 1; Condit et al. 2005). Similarly, the number of
plant families represented in forest communities varies from 47 in Korup, Cameroon, to 76 in Lambir, Malaysia (Ricklefs and Renner 2012).
Local dominance by one or a few species is found in primary rain forests
throughout the tropics. In the Asian tropics, the family Dipterocarpaceae (e.g.,
Dryobalanops aromatica) dominates, while many species in the leguminous family
Caesalpiniaceae dominate in the African and Neotropics (e.g., Gilbertiodendron
dewevrei in Congo, Mora excela in Trinidad, and Peltogyne gracilipes in Brazil).
A comprehensive assessment by Ter Steege et al. (2013) of the composition and
biogeography of tree communities from 1,170 inventory plots throughout Amazonia
yielded the stunning discovery that a mere 227 of the roughly 16,000 tree species in
this region account for half of the trees. Species of palm trees in the Arecaceae are
predominant, as well as species in the Myristicaceae, Lecythidaceae, and commonly
cultivated trees. Most of these so-called hyperdominant species forming predictable
oligarchies are only dominant in certain forest types and, while they demonstrate
large geographic ranges, show strong evidence of habitat specialization though a
broad range of shade tolerance is represented. It is the rare species, with average
abundances of 1 individual per hectare that drive species richness of tropical
communities (Table 2). The striking discovery that a small suite of species largely
drives Amazonias biogeochemical cycling opens areas of inquiry into the implications of species-specific effects of climate change on productivity and phenology in
this region. Elucidating mechanisms that promote dominance and monodominance
also provide important conceptual contrast to those explaining high species diversity.
260
R.E. Gallery
Table 1 Forest diversity by region from large tropical forest plots associated with the Center for
Tropical Forest Science (CTFS). Lines in the table denote Southeast Asian, Neotropical, and
African regions. Annual precipitation for each forest is shown in millimeters (mm), and the
number of dry season months is in parentheses. Two different size classes are shown for the full
plot and per hectare. Sites marked with an asterisk were < 25 ha, and data for those sites are based
on the full 16 or 20 ha. Main references for each plot are footnoted (Redrawn with permission
Condit et al. 2005)
Plot
size
(ha)
Lambia, Borneo, 52
Malaysiaa
Huai Kha
50
Khaeng,
Thailandb
Mudumalai,
50
Indiac
Pasoh,
50
Peninsular
Malaysiad
Sinharaja, Sri
25
Lanka
Palanan,
16
Philippines*
Barro Colorado, 50
Panamae
La Planada,
25
Colombia
Yasuni,
25
Ecuadorf
Luquillo, Puerto 16
Ricog*
Korup,
50
Cameroon
Ituri,
D.R. Congoh:
Lenda
20
(monodominant)
Edoro (mixed)
20
a
mm annual
precipitation (dry
season in mo.)
2,664 (0)
Species
per ha
10 cm
dbh
245.7
Species in
full plot
10 cm dbh
1,008
Species
per ha
1 cm
dbh
618.1
Species in
full plot
1 cm dbh
1,179
1,476 (6)
65.6
217
101.8
259
1,206 (6)
22.0
63
25.6
72
1,788 (0)
207.3
678
496.5
814
5,074 (0)
71.2
167
142.7
205
3,218 (4)
98.9
262
201.6
335
2,551 (3)
90.7
227
168.0
301
4,087 (0)
85.0
172
150.1
219
3,081 (0)
253.6
820
665.2
1,104
3,548 (0)
42.2
87
77.6
140
5,272 (3)
85.4
307
235.1
494
1,674 (2)
49.1
211
166.0
365
1,785 (2)
67.0
212
172.2
380
261
Table 2 Species rarity and dominance by region. Percent of rare species (those with 0.3
individuals per ha) at each of the plots and relative abundance of the dominant species. Both are
given as mean 95 % confidence limits, based on replicate 20-ha subquadrats. Confidence limits
for Congo sites could not be calculated, since the plots were only 20 ha; for sites marked with an
asterisk, the estimates are based on the full 16 ha and also lack confidence limits. Dominant
species for each site is listed along with authority and family (Redrawn with permission Condit
et al. 2005)
Plot
Lambia, Borneo, Malaysia
% Rare species
14.9 3.7
% Dominance
2.6 1.0
44.8 1.5
10.0 5.2
Mudumalai, India
41.7 4.8
22.8 6.5
19.2 3.5
2.7 0.3
16.6 0.9
12.1 0.4
Palanan, Philippines*
37.9
5.6
25.6 2.7
15.7 1.9
La Planada, Colombia
24.2 2.9
15.6 0.1
Yasuni, Ecuador
31.1 0.6
3.1 0.1
40.7
19.6
Korup, Cameroon
29.2 2.6
8.3 1.5
48.4
45.0
Edoro (mixed)
52.2
41.8
Dominant species
Dryobalanops aromatica
Gaertner (Dipterocarp-)
Croton oblongifolius
Roxb. (Euphorbi-)
Kydia calycina
Roxb. (Malv-)
Xerospermum noronhianum
Blume (Sapind-)
Humboldtia laurifolia
M. Vahl (Fab-)
Nephelium lappaceum
Poiret (Sapind-)
Hybanthus prunifolius
Schulze-Menz (Viol-)
Faramea caffeoides
C.M. Taylor (Rubi-)
Matisia oblongifolia
Poeppig & Endl. (Malv-)
Palicourea riparia
Benth. (Rubi-)
Phyllobotryum spathulatum
M
ull. Arg. (Salic-)
Scaphopetalum dewevrei
Wildem. & Th. Dur. (Malv-)
Scaphopetalum dewevrei
262
R.E. Gallery
competitive exclusion of species from the community. They maintain alpha diversity within communities by reducing interspecific competition or through densitydependent pest regulation of plant populations. Pervasive dispersal and recruitment
limitation, whereby a species does not successfully establish in all sites it is capable
of occupying, further reduce the extirpation of less competitive species in a
community.
Forest disturbances such as tree fall gaps create light and nutrient heterogeneity
that generate niche opportunities allowing tree species coexistence across the
continuum of light-demanding pioneer to longer lived, better defended shadetolerant species. Tree fall gaps are colonized in a number of ways that can alter
regeneration or successional pathways. Light-demanding pioneer species germinate
readily from soil seed banks when the high light quality and temperature conditions
from gaps arise. The rapid growth rate of these species results in a developing
understory that leads to favorable microsites for other species to recruit. Recruitment from the seedling bank is equally common. Shade-tolerant seedlings and
saplings persisting in the understory for decades are also able to exploit the high
light environment of gaps and respond with rapid growth rates. Vegetative propagation, clonal shoots, and lateral growth from vines and lianas are also pathways for
gap colonization, and plant recruitment and growth rates thin and slow as competition for light increases. Despite the importance of forest gaps, there is little
evidence that variations in adaptation to disturbance account for the high alpha
tree species diversity of tropical rain forests. Disturbance is nevertheless one of
several factors that add to seemingly unpredictable microclimatic conditions within
tropical forests.
263
thought to regulate populations (see Janzen 1970; Connell 1971), however, does not
seem to hold either for plantphytophage or plantpathogen interactions in tropical
forests. Polyphagy in insects (Novotny et al. 2002) and fungi (Gilbert and Webb
2007) is the more common strategy in species-rich communities with high numbers
of locally rare species. Nevertheless, plant preferences of pests and the variation in
plant responses to common pests appear to be sufficient to facilitate coexistence
among plants as described in the JanzenConnell hypothesis.
264
R.E. Gallery
35
High estimate
Low estimate
30
25
20
15
10
5
0
0
35
30
25
y = 0.03x 2 1.0x + 16.5
R 2 = 0.30
20
15
10
5
0
10
12
14
16
18
20
22
24
26
28
30
35
30
25
20
15
10
5
0
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
Fig. 5 The relationships between low and high estimates of NPP for 39 old-growth tropical forest
sites around the world and (a) annual precipitation (P), (b) mean annual temperature (T ), and (c)
the ratio T/P 100 (Reprinted with permission Clark et al. 2001)
265
where soils are the dominant C store. Although there are seasonal patterns of plant
growth in the tropics, high solar radiation and a relatively stable warm, wet climate
provide more consistently suitable conditions for growth than drier and colder
regions. Consequently, tropical forests account for approximately 40 % of NPP.
An estimated 60 % of tropical forests are classified as secondary or degraded forests
(Chazdon 2003), meaning tropical deforestation has considerable implications for
Earths carbon cycle.
There is evidence that aboveground biomass production is increasing in the
forests of South America, Africa, and Asia, though notably not Australia. The
primary mechanisms driving this trend are thought to include increased resource
availability through the effect of rising atmospheric CO2, air temperature, and solar
radiation on NPP, and forest recovery from past disturbances. The contrasting
pattern in Australian tropical rain forests is linked to the magnitude, frequency,
and scale of natural disturbances such as cyclones and strong droughts from El Nino
events. Intact tropical forests are net C sinks, but the uptake of C (1.1 0.3 Pg C
year1) in intact tropical forests is counteracted by the emissions from tropical
biome conversion a net C source to the atmosphere of 1.3 0.2 Pg C year1 that
results in a tropical biome net C balance of approximately zero (Malhi 2010).
However, there are few studies under ambient or elevated CO2 conditions where
the net C uptake of tropical forests has been quantified, and the role of tropical
forests in Earths C cycle, while critical, is far from understood.
266
R.E. Gallery
Table 3 Approximate geographic extent of contemporary forest cover, deforestation, and selective logging by region in the humid tropical forest biome. Values are in km2, with percentage of
biome extent also givena (Redrawn with permission Asner et al. 2009)
Region
Africa
Asia/
Oceania
Central
America/
Caribbean
South
America
Total
Total
biome
extent
(km2)
2,918,511
7,191,529
685,840
8,826,966
3,194,632
(36.2 %)
19,622,846 10,016,282
(51.0 %)
5,632,334
(63.8 %)
9,606,564
(49.0 %)
Forest area
cleared
20002005c
(km2)
14,972 (0.5 %)
Selective
loggingd
(2000s)
(km2)
561,153
(19.2 %)
93,955 (1.3 %) 1,777,963
(27.2 %)
9,687 (1.4 %) 36,097
(5.3 %)
156,001
(1.8 %)
274,615
(1.4 %)
1,603,166
(18.2 %)
3,978,379
(20.3 %)
Percentage of regional biome extent is in parentheses, except in the column totals (last row),
where percent refers to the global biome extent. Differences in the composition, spatial extent,
temporal scale, and quality of the available data make it difficult to quantitatively compare rates of
deforestation and selective logging. They are listed here to provide a general global perspective on
the magnitude of reported or detected contemporary changes among these land-use processes
b
Forest cover in 2005 calculated as 2000 forest cover minus losses from 2000 to 2005 with data
from Hansen et al. (2008). Percent forest cover is based on percent within each 500 m grid cell,
followed by conversion to vector format for global calculations
c
Calculated from Hansen et al. (2008)
d
Logging does not represent actual harvested trees, but rather regional forest areas in which timber
operations occur
manipulated landscapes are susceptible to small island effects such as the loss of
species diversity through unsustainable coexistence in shrinking patches, loss of
population genetic diversity through restricted migration and shrinking population
size, and nonnative and disturbance-adapted species invasions that alter community
diversity and successional pathways. Strong edge effects in forest patches increase
tree mortality of drought-sensitive species and from physical exposure to increased
winds that cause blow down. Deposition of dust and aerosols rich in N and P from
surrounding agriculture and development alters plant growth rates. Increased evaporation, decreased soil moisture, and the accumulation of litter increase susceptibility
to fires, and, indeed, contemporary fire occurrence in tropical forests is largely
associated with forest edges (Cochrane 2003). Nevertheless, these forest mosaics
are the future of tropical regions, and thoughtful management can benefit agriculture
as well as preserve forests and their ecosystem services that contribute to water quality
and global food supply (e.g., pollinators).
Stronger ENSO effects are increasing the frequency and severity of droughts,
fires, hurricanes and cyclones, and flooding events. Historical records and charcoal
in soil profiles show that tropical forest fires, even in wetter forests, are not
unprecedented. Fire is considered endemic but rare in most tropical rain forests,
267
Future Directions
The tropical biome is undergoing significant change. Understanding the drivers and
impacts of these changes will require sustained advances across multiple disciplines.
Ultimately as a society, we are left asking what is the capacity of our remaining and
268
R.E. Gallery
China
Philippines
Thailand
ecoregion
endemics
25 - 50
51 - 100
101 - 169
b
Malaysia
oil palm
area (%)
0
0-1
1-2
2-5
5 - 10
10 - 20
20 - 40
suitable for
oil palm
no forest
forest
oil palm
area (%)
Indonesia
0
0-1
1-2
2-5
5 - 10
10 - 20
20 - 40
TRENDS in Ecology & Evolution
Fig. 6 Global distribution of oil palm and potential conflicts with biodiversity: (a) areas of highest
terrestrial vertebrate endemism (ecoregions with 25 or more endemics are shown), (b) global
distribution of oil palm cultivation (harvested area as percentage of country area), (c) agriculturally suitable areas for oil palm (with and without forest), and (d) oil palm-harvested area in
Southeast Asia. In (b) and (d), Brazil, Indonesia, Malaysia, the Philippines, and Thailand are
subdivided by province, but other countries are not. Data are for 2006, except for the Philippines
and Thailand, where 2004 data are the most recent available (Sources: (a) World Wildlife Fund
(2006) WildFinder: online database of species distributions, version Jan-06, http://www.
worldwildlife.org/wildfinder; (b, d) world: http://faostat.fao.org; Brazil: http://www.ibge.gov.br/
estadosat; Indonesia: http://www.deptan.go.id; Malaysia: http://econ.mpob.gov.my/economy/
annual/stat2006/Area1.7.htm; Philippines: http://www.bas.gov.ph/downloads_view.php?id127;
Thailand: http://www.oae.go.th/statistic/yearbook47/indexe.html; (c) forest area: European Commission Joint Research Centre (2003) Global Land Cover 2000 database, http://www-gem.jrc.it/
glc2000; oil palm suitability: updated map from G. Fischer, first published in Fischer,
G. et al. (2002) Global Agro-Ecological Assessment for Agriculture in the 21st Century: Methodology and Results, International Institute for Applied Systems Analysis and Food and Agriculture Organization of the United Nations) (Reprinted with permission Fitzherbert et al. 2008)
269
References
Asner GP, Rudel TK, Aide TM, Defries R, Emerson R. A contemporary assessment of change in
humid tropical forests. Conserv Biol. 2009;23(6):138695.
Asner GP, Mascaro J, Anderson C, Knapp DE, Martin RE, Kennedy-Bowdoin T, van Breugel M,
Davies S, Hall JS, Muller-Landau HC, Potvin C, Sousa W, Wright J, Bermingham E. Highfidelity national carbon mapping for resource management and REDD+. Carb Balance Manag.
2013;8(1):114.
Bagchi R, Gallery RE, Gripenberg S, Gurr SJ, Narayan L, Addis CE, Freckleton RP, Lewis
OT. Pathogens and insect herbivores drive rainforest plant diversity and composition. Nature.
2014;506(7486):8588.
Bunyavejchewin S, Baker PJ, LaFrankie JV, Ashton PS. Stand structure of a seasonal dry
evergreen forest at Huai Kha Khaeng Wildlife Sanctuary, Western Thailand. Nat Hist Bull
Siam Soc. 2001;49:89106.
Chazdon RL. Tropical forest recovery: legacies of human impact and natural disturbances.
Perspect Plant Ecol Evol Syst. 2003;6:5171.
Clark DA, Brown S, Kicklighter DW, Chambers JQ, Thomlinson JR, Ni J, Holland EA. Net
primary production in tropical forests: an evaluation and synthesis of existing field data. Ecol
Appl. 2001;11(2):37184.
Cleveland CC, Townsend AR, Taylor P, Alvarez-Clare S, Bustamante M, Chuyong G, Dobrowski
SZ, Grierson P, Harms KE, Houlton BZ, Marklein A, Parton W, Porder S, Reed SC, Sierra CA,
Silver WL, Tanner EVJ, Wieder WR. Relationships among net primary productivity, nutrients
and climate in tropical rain forest: a pan-tropical analysis. Ecol Lett. 2011;14(9):93947.
Cochrane MA. Fire science for rainforests. Nature. 2003;421(6926):9139.
Condit R, Ashton P, Balslev H, Brokaw N, Bunyavejchewin S, Chuyong G, Co L, Dattaraja HS,
Davies S, Esufali S, Ewango CEN, Foster R, Gunatillek N, Gunatillek S, Hernandez C,
270
R.E. Gallery
271
Dallmeier F, Comiskey JA, editors. Forest biodiversity diversity research, monitoring, and
modeling. Paris: UNESCO, the Parthenon Publishing Group; 1998. p. 42946.
Malhi Y. The carbon balance of tropical forest regions, 19902005. Curr Opin Environ Sustain.
2010;2(4):23744.
Manokaran N, LaFrankie JV, Kochummen KM, Quah ES, Klahn J, Ashton PS, Hubbell SP. Stand
table and distribution of species in the 50-ha research plot at Pasoh Forest Reserve. Kepong,
Malaysia: Forest Research Institute of Malaysia; 1992.
Myers N, Mittermeier RA, Mittermeier CG, Da Fonseca GA, Kent J. Biodiversity hotspots for
conservation priorities. Nature. 2000;403(6772):8538.
Novotny V, Basset Y, Miller SE, Weiblen GD, Bremer B, Cizek L, Drozd P. Low host specificity
of herbivorous insects in a tropical forest. Nature. 2002;416(6883):8414.
Ricklefs RE, Renner SS. Global correlations in tropical tree species richness and abundance reject
neutrality. Science. 2012;335(6067):4647.
Romoleroux K, Foster R, Valencia R, Condit R, Balslev H, Losos E. Especies lenosas (dap >1 cm)
encontradas en dos hectreas de un bosque de la Amazona ecuatoriana. In: Valencia R,
a de Plantas. Quito: Pontificia
Balslev H, editors. Estudios Sobre Diversidad y EcologU
Universidad Catolica del Ecuador; 1997. p. 189215.
Schongart J, Junk WJ, Piedade MTF, Ayres JM, Huttermann A, Worbes M. Teleconnection
between tree growth in the Amazonian floodplains and the El Nino-Southern Oscillation effect.
Glob Chang Biol. 2004;10:68392. doi:10.1111/j.1529-8817.2003.00754.x.
Sodhi NS, Brook BW, Bradshaw CJ. Tropical conservation biology. Oxford, UK: Blackwell;
2007.
Sukumar R, Dattaraja HS, Suresh HS, Radhakrishnan J, Vasudeva R, Nirmala S, Joshi
NV. Longterm monitoring of vegetation in a tropical deciduous forest in Mudumalai, southern
India. Curr Sci. 1992;62:60816.
ter Steege H, Pitman NC, Sabatier D, Baraloto C, Salomao RP, Guevara JE, Phillips OL,
et al. Hyperdominance in the Amazonian tree flora. Science. 2013;342(6156):1243092.
doi:10.1126/science.1243092.
Thompson J, Brokaw N, Zimmerman JK, Waide RB, Everham III EM, Lodge DJ, Taylor CM,
Garcia-Montel D, Fluet M. Land use history, environment, and tree composition in a tropical
forest. Ecol Appl. 2002;12:134463.
Townsend AR, Asner GP, Cleveland CC. The biogeochemical heterogeneity of tropical forests.
Trends Ecol Evol. 2008;23(8):42431.
ndez C, Romoleroux K, Losos E,
Valencia R, Foster RB, Villa G, Condit R, Svenning JC, HernA
MagIrd E, Balslev H. Tree species distributions and local habitat variation in the Amazon: a
large forest plot in eastern Ecuador. J Ecol. 2004;92:21429.
van der Heijden GM, Schnitzer SA, Powers JS, Phillips OL. Liana impacts on carbon cycling,
storage and sequestration in tropical forests. Biotropica. 2013;45:68292.
van Schaik CP, Terborgh JW, Wright SJ. The phenology of tropical forests: adaptive significance
and consequences for primary consumers. Annu Rev Ecol Syst. 1993;24:35377.
Worbes M. One hundred years of tree ring research in the tropics a brief history and an outlook to
future challenges. Dendrochronologia. 2002;20:21731.
Zimmerman JK, Everham EMI, Waide RB, Lodge DJ, Taylor CM, Brokaws NVL. Responses of
tree species to hurricane winds in subtropical wet forest in Puerto Rico: implications for
tropical tree life histories. J Ecol. 1994;82:91122.
Further Reading
Cernusak LA, Winter K, Dalling JW, Holtum JA, Jaramillo C, Korner C, Leakey ADB, Norby RJ,
Poulter B, Turner BL, Wright SJ. Tropical forest responses to increasing atmospheric CO2:
current knowledge and opportunities for future research. Funct Plant Biol. 2013;40:53151.
272
R.E. Gallery
Hubbell SP. Tropical rain forest conservation and the twin challenges of diversity and rarity. Ecol
Evol. 2013;3(10):326374.
Laurance WF, Sayer J, Cassman KG. Agricultural expansion and its impacts on tropical nature.
Trends Ecol Evol. 2014;29(2):10716.
Lewis SL, Lloyd J, Sitch S, Mitchard ETA, Laurence WF. Changing ecology of tropical forests:
evidence and drivers. Annu Rev Ecol Evol Syst. 2009;40:52949.
Molbo D, Machado CA, Sevenster JG, Keller L, Herre EA. Cryptic species of fig-pollinating
wasps: implications for the evolution of the figwasp mutualism, sex allocation, and precision
of adaptation. Proc Natl Acad Sci. 2003;100(10):586772.
Pitman NC, Terborgh JW, Silman MR, Nunez VP, Neill DA, Ceron CE, Palacios WA, Aulestia
M. Dominance and distribution of tree species in upper Amazonian terra firme forests.
Ecology. 2001;82(8):210117.
Schemske DW, Mittelbach GG, Cornell HV, Sobel JM, Roy K. Is there a latitudinal gradient in the
importance of biotic interactions? Annu Rev Ecol Evol Syst. 2009;40:24569.
Wright SJ. The future of tropical forests. Ann N Y Acad Sci. 2010;1195(1):127.
10
Russell K. Monson
Contents
What Is a Forest? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
The Climate and Phytogeography of Temperate Forests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
The Geologic Origins of Temperate Forests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Temperate Forests and the Concept of Ecological Succession . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Temperate Forest Carbon Cycling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Temperate Forest Net Primary Productivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Temperate Forest Nitrogen and Phosphorus Cycling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Temperate Forest Water Cycling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Plant Functional Traits in Temperate Forest Trees . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Temperate Forests and Disturbance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Future Directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
275
275
277
278
281
282
284
285
287
289
294
295
Abstract
273
274
R.K. Monson
the exceptionally high C:N ratio of wood (often >300). As a result, temperate
forests are capable of storing high quantities of carbon that are assimilated
from the reservoir of atmospheric CO2.
Classic, historical concepts in ecology, such as succession, have been developed from studies of temperate forest ecosystems. Forest succession refers to
decadal-scale transitions in community composition. Each shift in community
composition causes changes in the forest microenvironment, which in turn
causes further changes in community composition. Traditionally, this pattern
of progressive change in community composition and associated feedbacks to
forest microenvironment was viewed within a highly deterministic framework.
More recently, ecological concepts, such as gap theory, have emerged from
the older concepts of succession and have been developed with greater emphasis on stochasticity. Both succession and gap theory has contributed greatly to
our understanding of the causes of natural and anthropogenic changes to the
species composition of temperate forest ecosystems.
Nitrogen and phosphorus (N and P) are cycled through temperate forest
ecosystems through a process of coupled recycling involving serial relationships between plants and soil microorganisms. N or P that is deposited to the
soil through litter production is transformed from organic to inorganic forms
through microbial mineralization, producing nitrate and phosphate ions,
which can then be re-assimilated by plants and used to construct new organic
biomass. Leaching of phosphate and nitrate from forest soils (especially
nitrate in temperate forests) prior to re-assimilation by plants represents an
important nutrient loss process and often limits forest biomass production.
Root-fungal symbioses, called mycorrhizae, are well developed in temperate
forest ecosystems. The hyphal biomass from the fungus radiates from associated roots and increases the capacity for trees to capture nitrate and phosphate
prior to leaching and, in some cases, allows trees to take up organic nitrogen
(such as small proteins or single molecules of amino acids). The acquisition of
organic forms of nitrogen (and to some extent phosphorus) short-circuits the
conventional form of biogeochemical cycles (alternating between plants and
microbes) and increases the efficiency of nutrient retention in the ecosystem.
Most water that is cycled through forests is used to sustain a favorable energy
balance. Evapotranspiration from forests facilitates the loss of heat that is
absorbed as net radiation (from the sun and sky) and returns water to the
atmosphere, thus sustaining the terrestrial water cycle.
Trees in temperate forests (especially in North America and Europe) have been
exposed to increasing physiological stress in recent decades due to the increased
frequency of drought and high temperatures. These stresses have the potential to
reduce forest growth and may be responsible for the observed weakening of
forest carbon sinks globally. Climate-induced stress, in turn, exposes temperate
forests to an increased frequency of epidemic insect outbreaks and associated
high rates of herbivory, as well as shorter fire return cycles. The combination of
abiotic and biotic stress is likely responsible for an increase in observed mass
tree mortality in temperate forests of the Northern Hemisphere.
10
275
What Is a Forest?
The term forest has been used since at least the Middle Ages, when William the
Conquerer consolidated much of the knowledge about his newly acquired lands in
the Domesday Book of 1086 CE (Common Era). Royal forests, as listed in the
Domesday Book, referred to unbounded lands intended to raise wild animals that
could be hunted by the monarch and other members of the royal family. Forests
were not classified according to ecological or botanical attributes, but rather as legal
entities afforded protection by laws and management. Forests at this time included
grasslands, woodlands, heathlands, and even agricultural fields.
In more recent times, the term forest has been associated with woodlands, and
most dictionary definitions include reference to a high density of trees. The US
National Vegetation Classification Scheme, which is produced through oversight
by the US Federal Geographic Data Committee (an interagency committee led by
representatives of the US Geological Survey), distinguishes forests from woodlands. Forests are areas with trees forming overlapping crowns with 60100 %
areal coverage. Woodlands are more open, with 2560 % crown coverage. Even
with these rather precise definitions, however, ecologists will often use the term
more loosely, for example, in describing the great kelp forests in the coastal
oceans of temperate and polar regions.
276
R.K. Monson
Summers tend to be warm but moist enough to support the relatively high water
demands of the tree life-form, and winters tend be cool to cold, but also moist.
At higher elevations, winter moisture is deposited as snow. Limited access to
liquid water and cold winter temperatures force trees in higher-elevation
temperate forests to minimize metabolic activity until the time of spring or summer
snowmelt. Even evergreen, coniferous trees in these ecosystems tend to
downregulate their metabolic activities in the winter to levels just sufficient to
sustain basal respiration. Thus, in high-elevation, temperate coniferous forests,
winter forest carbon balances are characterized by deficits, causing the forests
to be net carbon sources, at least seasonally. In lower-elevation coastal forests
winters are cool to cold, but moisture continues to be deposited as rain. In these
coastal rainforests, evergreen trees remain metabolically during the winter, and
wintertime photosynthesis can represent a significant fraction of annual net primary
productivity.
Temperate forests include broadleaf and needleleaf tree forms. Temperate
forest trees tend to have long generation times, lasting multiple decades,
compared to plants in other mid-latitude ecosystems (though temperate alpine
plants can also have multi-decadal life spans). As a result, forests tend to
migrate slowly across the landscape in response to climate changes. This creates
disequilibrium between climate change and forest distribution. Emerging from the
Last Glacial Maximum, and midway into the Holocene Era, poleward forest
migrations in the Northern Hemisphere have been estimated as 22.5 km yr1
(based on pollen records; Davis 1989), which is considerably slower than the
617 km yr1 estimated for non-tree species during the current, Anthropocene
warming (Parmesan and Yohe 2003; Chen et al. 2011). The slow migration of
temperate forests in response to a rapidly changing climate poses interesting
questions as to how temperate forest ecosystems will adjust to future humaninfluenced climate regimes.
Forests tend to have a unique nutrient stoichiometry, particularly with regard to
C:N ratios; because of the high C:N ratio of wood (often 300 or higher), whole-tree
C:N ratios (wood, leaves, and roots) tend to be between 100 and 200 for temperate
forest trees (Norby et al. 1999). Herbaceous plants often exhibit C:N ratios that are
less than 100 and often less than 50 (Lebreton and Gallet 2007). Because of their
high C:N ratios, temperate forests account for much of the net carbon dioxide
assimilated from the atmosphere during each years growing season. Based on the
Food and Agriculture Organization of the United Nations, as of the year 2000 CE,
temperate broadleaf forests and temperate evergreen forests covered approximately
400 million and 100 million ha of the Earths surface, respectively. Taking into
account all temperate forests on the globe, it is estimated that net primary productivity for this biome type is ~8 Pg C yr1 (Saugier et al. 2001). This rate of carbon
uptake is approximately 13 % of the total global rate of photosynthetic carbon
sequestration and similar to the total annual anthropogenic CO2 emissions due to
fossil fuel combustion. Thus, temperate forest ecosystems represent a vital component of the Earths carbon budget and must take a central role in discussions of
global carbon cycle management.
10
277
Fig. 1 Proposed landscape for Devonian temperate forests 385 mya. The forest is composed of
cycads, tree ferns, and Aneurophytaleans and is likely to be one of the oldest temperate forests yet
discovered. This landscape drawing was based on fossils uncovered in Schoharie County,
New York, and the drawing was produced by Frank Mannolini of the New York State Museum,
Albany, New York (Reproduced here with their kind permission. Copyright remains with the
New York State Museum)
278
R.K. Monson
The title of Thoreaus essay was The Succession of Forest Trees. Thoreaus
essay focused on the role of animals as they move among plant community types in
establishing the seed bank for future changes in forest community composition.
During the early part of the twentieth century, Frederic Clements and Henry
Gleason debated openly about the nature of plant community changes and in
particular forest succession. Clements viewed ecological succession as deterministic, a time-dependent process of species replacements, responding to changes in
forest microclimate and soil fertility, which ended in the so-called climax community. Successional sequences could occur on newly developed substrates as they
were mineralized, such as volcanic ash or rock (primary succession), or they could
be reset following disturbances to established communities, such as following a
stand-replacing fire, or clear-cut logging (secondary succession). In either case, the
climax community could be predicted through observation of other stable communities in the same climate and geographic regimes. The climax state was
predominantly controlled by climate, soil fertility, and their interactions with the
10
279
280
R.K. Monson
Fig. 2 Forest succession patterns predicted for a native beech-dominated forest near Davos,
Switzerland, as provided by the forest dynamics model, ForClim, ver. 2.9. The pattern demonstrates the expected shifts in community composition during secondary succession, with the
deciduous coniferous species, Larix decidua, emerging as a pioneer species, and giving way to
eventual dominance by Abies alba (white fir) and Fagus sylvatica (common beech). The model
simulation begins with the current climate at Year 0 and progresses through a series of future
climate scenarios for a period of one-an-a-half millennia (Redrawn from Bugmann (2001))
10
281
282
R.K. Monson
(i.e., the products of new GPP). The latter component of Re is often partitioned
further into the CO2 efflux from aboveground plant tissues (often included as a
component of net primary productivity, NPP) and that from roots and soil microorganisms that are symbiotically associated with roots (e.g., mycorrhizal fungi and
rhizospheric bacteria). It is difficult to clearly distinguish the dependencies of soil
respiration on recent GPP versus older soil organic matter, as respiratory substrates
exist along a continuum of ages. In some studies, the components have been
distinguished through the experimental girdling of all trees in a forest stand or
through large-scale labeling of photosynthetic products in forest trees using isotope
tracers (Hogberg and Read 2006). Tree girdling essentially chokes the flow of
photosynthetic products from the leaves (or needles) to the soil, thus eliminating
the soil respiration component linked to recent photosynthetic activity. Labeling of
the forest with 13CO2 has been accomplished with giant tents to contain the applied
label, followed by time-dependent tracing of the paths taken by labeled photosynthetic compounds and the kinetics by which 13CO2 is released from the labeled
photosynthetic products through the processes of plant and microbial respiration. In
the final accounting, all CO2 released by ecosystem respiration is dependent on the
rate of GPP, as this determines the rate by which carbon substrates enter the
ecosystem and are used by plant or microbial cells as respiratory energy sources.
The rates, at which these substrates are utilized, however, are subject to modification according to abiotic factors, such as temperature and moisture availability.
The net carbon uptake of forest ecosystems is typically measured using towers
that extend above the canopy and have instruments attached that are capable of
measuring the statistical covariance between the vertical wind speed (up versus
down) and the CO2 concentration in the atmosphere near the canopy surface. This is
the so-called eddy covariance (or eddy flux) approach (Fig. 3). Using this approach,
a continuous record of the cumulative rate of biological carbon sequestration can be
measured for the forest, including all carbon stored in the trees and soil. Using this
approach, one can study the effect of climate variation on forest carbon uptake
across long (decadal) time scales. It is through this type of study, combined with
computer models of ecosystem processes, that insight is being gained into the
feedbacks between climate change and forest carbon uptake.
10
283
Fig. 3 Left panel. A 12-year record of net ecosystem productivity (NEP) measured in a subalpine
forest in Colorado, USA. The sawtooth record shows seasonal variation in the cumulative NEP,
with decreases in NEP shown for winter (the forest continues to respire but GPP is near zero) and
increases in NEP shown for the growing season (GPP exceeds Re resulting in forest carbon
sequestration). In the lower left panel, the annual sum of forest carbon sequestration is shown
for the 12-year time series. Right panel. A picture of the flux tower used to measure NEP from the
Niwot Ridge subalpine forest. Instruments near the top of the tower record the turbulent fluxes of
CO2, and a profile of mean CO2 concentration measurements is made along the length of the tower
to account for CO2 that is retained within the canopy below the turbulent flux instruments. This
technology is often referred to as recording the eddy flux and CO2 storage flux, respectively,
and the sum of these values provides an estimate of NEP
284
R.K. Monson
Overall, temperate forests represent relatively large carbon sinks. Even compared to tropical forests, temperate forests can store large quantities of carbon.
Past estimates of tropical forest net primary forest productivity have been highly
variable among sites and years but have generally fallen within the range 0.22 kg
m2 yr1 (Clark et al. 2001). Thus, while temperate forest productivity is within the
lower range of that estimated for tropical forests, it can be as high as 50 %,
depending on the forest site and year of consideration.
10
285
This mutual exchange allows both the tree and fungus to sustain favorable C:N
ratios. However, the relationship between roots and fungi in the mycorrhizal
exchange of C and N is mitigated to some extent by the availability of soil N and
the requirements for fungi to immobilize a required minimal fraction. When soil N is
relatively more abundant, fungal transfer of organic N to tree roots is favored, but
when soil N is relatively less abundant, immobilization by mycorrhizal fungal
hyphae is favored. Discovery of the use of organic N by both plants and soil
microorganisms has changed our conventional view of forest nitrogen cycles.
Whereas plant N nutrition was once viewed as being dependent on microbial
mineralization, it is now recognized that plants and microorganisms interact directly,
in both symbiotic and competitive ways, to partition organic soil N. These interactions will continue to be the focus of forest N cycling for years to come.
Ignoring the obvious constraints by low soil temperatures and limitations to
microbial biomass, conventional wisdom holds that the maximum rate of primary
production in temperate forests is ultimately limited by the availability of soil N
(Reich et al. 1997). Other factors, such as soil water limitations, high air temperatures, cloudy weather, and low atmospheric humidity, can limit rates of productivity
over short time scales, but over longer, decadal time scales, soil N availability will
set a clear upper limit on NPP. The constraint of soil N limitation has placed selective
pressures on temperate forest trees to evolve efficient rates of N recycling between
the soil and plant and to retain and store N in plant tissues prior to seasonal or multiseasonal leaf senescence. The question as to what, in turn, determines the rate of soil
N availability has been addressed in conceptual models (Vitousek and Field 1999).
One prominent limitation that has the potential to determine rates of N fixation in
terrestrial ecosystem (and aquatic ecosystems, for that matter) is P availability.
N-fixing organisms are often limited by P, and P weathers from parent minerals at
relatively slow rates. Thus, a cascade of controls can be proposed for the long-term
limitation of temperate forest productivity, extending from P limitations to N fixing
pioneer species, N limitations to later successional species, and ultimately light
limitations to carbon fixation rates as canopies close. The role of multiple nutrient
constraints in limiting primary productivity is demonstrated in the meta-analysis of
over 200 fertilization studies in temperate deciduous forests conducted by
Vadeboncoeur (2010) (Fig. 4). Most of the studies included in that analysis showed
positive growth responses to the addition of phosphorus, calcium, and nitrogen.
These types of results indicate that the nutrient limitations to primary productivity in
forests are complex in their nature and interactions. It may be too simple to rely on
statements in the conventional wisdom, such as nitrogen limits productivity in
temperate forests and phosphorus limits productivity in tropical forests.
286
R.K. Monson
Fig. 4 Frequency of
response ratios for a metaanalysis of 208 fertilization
studies of North American
deciduous forests. The
response ratio is the ratio of
net primary production in the
presence of an experimental
fertilization treatment relative
to control plots with no
treatment. The vertical line
indicates a response ratio of
1.0 for reference (Redrawn
from Vadeboncoeur (2010))
surface and transpiration from leaves (and branches to a lesser extent). The combined evapotranspiration from a forest is driven by energy inputs and biological
attributes of the vegetation a truly biophysical process. Energy that is absorbed
from the sun and atmosphere by a forest must be partitioned into various energy loss
processes, or it will contribute to an increase in forest surface temperatures. This is
the nature of thermodynamics and the requirement for conservation of energy. The
energy loss mechanisms that are available to forests include radiative heat loss
(according to Kirchhoffs law), sensible heat loss (through conduction of heat to the
atmosphere or deeper soil layers), or latent heat loss (through evapotranspiration).
The loss (through photosynthesis) or gain (through respiration) of energy is minimal compared to the processes of reradiation, sensible and latent heat loss. The
tendencies for a forest to lose heat through sensible or latent transfers are to some
extent, mutually exclusive. As a forest loses latent heat, its surfaces will cool, which
in turn reduces the capacity for radiative heat loss and, assuming that canopy
surfaces are warmer than the atmosphere, sensible heat loss. As a forest canopy
conducts sensible heat to the atmosphere, less energy will be available to drive
latent heat loss, and radiative heat loss will decrease. During periods of ample soil
moisture, evapotranspiration rates are likely to be highest, and sensible heat losses
will concomitantly decrease in importance. In contrast, during periods of drought,
latent heat losses will be more limited, and the importance of sensible heat loss is
likely to increase.
Temperate forest canopies, especially at high elevations, can have significant
influences on ecosystem hydrology and the delivery of water resources to the
watersheds that support human communities. Forest canopies intercept and retain
a significant fraction of rain, especially during events with smaller rain drops and
lesser drop velocity. In those cases, intercepted precipitation can be directed back to
the atmosphere through evaporation from leaf surfaces, thus reducing canopy
throughfall and decreasing delivery to the soil. During the winter, canopies reduce
snowpack depth within the forest, thus storing less water for subsequent melt and
10
287
288
R.K. Monson
from temperate forest tree species. The formulation of PFTs is founded on the
recognition that natural selection works within populations of plants to produce
convergent and predictable patterns in form and function. Because many attributes
are linked in their effect on fitness, evolutionary modification of one attribute is
likely to cause a change in the fitness value of a second attribute, and these coupled
influences are likely to vary depending on environmental and growth habit context.
As an example, tree species that exhibit the evergreen growth habit tend to have
longer-lived leaves with lower metabolic rates, compared to species that exhibit the
deciduous growth habit. The concept of PFTs has also been extended to leaves
within a single tree; shade leaves tend to have longer life spans, lower N concentrations, and lower rates of metabolism, compared to sun leaves. A key area of research
that is currently underway is to separate the effects of genetics versus environment
on the patterns of coupled trait influences in plants from numerous types of biomes.
In the same way that aboveground suites of plant functional traits have been
recognized as constraining physiological function in predictable ways, belowground traits have been recognized recently as predictable predictors of biogeochemical, ecosystem processes (Phillips et al. 2013). As an example, lets return to
the topic of mycorrhizal associations between roots and fungi. Mycorrhizal symbioses are often classified as arbuscular mycorrhizal (AM) or ectomycorrhizal
(ECM). Arbuscular mycorrhizae involve the penetration of fungal hyphae into
cortical root cells, where the hyphae become highly branched and induce reorganization of the organelles and cytoskeleton of the cell in ways that enhance the
potential for reciprocal exchanges of carbohydrates and inorganic nutrients.
Ectomychorrhizal hyphae do not penetrate root cells, but instead form a dense
network between the epidermis and cortex, which then extends out into the soil. On
the outside surfaces of the roots, ECM hyphae can form a dense coat, or covering,
called the mantle, which exists at a higher biomass per unit of root length than the
hyphae of AM. Furthermore, ECM fungi are capable of exuding exoenzymes to the
soil, which catalyze reactions capable of mineralizing organic compounds from soil
litter and producing inorganic ions as well as small organic compounds capable of
resorption by the plant. AM fungi tend to not exude such enzymes and instead rely on
the inorganic ions secreted by decomposer bacteria essentially functioning in a
manner similar to fine roots. These different mycorrhizal associations represent
belowground plant functional traits and cause forest stands dominated by one type
or the other to exhibit distinct patterns of carbon and nitrogen cycling. Trees
associated with AM fungi tend to also produce leaf litter that decomposes rapidly,
releasing mineralized ions rapidly for fungus and plant resorption. Trees associated
with ECM tend to produce leaf litter that decomposes more slowly. With their suite of
exuded exoenzymes, ECM fungi are capable of obtaining their ions and organic
compounds from older, more recalcitrant, soil organic matter. Thus, in ECM trees
rapid litter decomposition rates are not as crucial to sustaining a favorable tree
nutrient balance. This framework, whereby belowground and aboveground traits
are constrained by common biogeochemical constraints, may be useful to extending
the concept of plant functional traits in ways that couple atmospheric and soil
processes in ecosystems and predict specific biogeochemical patterns and processes.
10
289
290
R.K. Monson
The need for suppression was progressively reinforced through the great fires of
1910 that burned over 1.2 million ha in the Western USA. By the late 1950s,
evidence began to accumulate, especially through research at the Southern Forest
Service Fire Laboratory in Macon, Georgia, that fire suppression actually increased
the threat of catastrophic crown fires. In other words, the risk of losing forests as a
natural resource was greater when fire suppression was practiced, than when it was
not. Fire suppression causes the accumulation of understory fuels that facilitate the
transformation of ground fires to stand-replacing, crown fires. Devastating fires in
1988 in Yellowstone National Park catalyzed creation of the US National Fire Plan,
which replaced fire suppression as the national fire strategy, with a different
strategy that emphasized canopy thinning and prescribed burns.
Recent studies using tree ring width and fire scars to reconstruct past histories of
fire frequency and its relation to climate have revealed that oscillatory climate
modes in the Earth system, particularly those associated with sea surface temperature, have a primary influence on fire frequency (Kitzberger et al. 2007). Patterns
of sea surface temperature oscillations are complex and variable among geographic
regions. However, one of the clear patterns that emerged is that in the Western USA,
years of warmer-than-normal sea surface temperature in the Tropical Pacific Ocean
(El Nino Southern Oscillations) often causes wet winters that facilitate an increase in
fuel load. Often, El Nino years are followed by years with cooler-than-normal sea
surface temperature in the Tropical Pacific (La Nina Southern Oscillations), which
produce drier-than-normal winters and summers. Dry weather during la Nina years
can increase flammability of the high fuel loads produced during the preceding El
Nino year. Longer-term modes in sea surface temperature, such as the Pacific
Decadal Oscillation (PDO) and Atlantic Multidecadal Oscillation (AMO), can interact with El Nino and La Nina oscillations to affect wildfire synchrony and frequency.
These connections between the Earths climate system and forest wildfires have been
best studied in montane coniferous forests of the Western USA. Tree growth in these
forests has decreased significantly during the period 19792008, compared to mean
changes in tree growth between 1896 and 2008 (Williams et al. 2010). The trend of
reduced forest growth was positively correlated with increased temperature and
increases in the frequency of negative (drought) precipitation anomalies during this
same period. As temperatures continue to rise and droughts become more frequent,
and extreme climate modes oscillate more frequently, these montane forests are
likely to experience even further decreases in growth.
Past studies on fossil plants have revealed that in general, during past geologic
eras of elevated CO2, such as during the middle Eocence (approximately 45 mya)
when atmospheric CO2 concentrations were in the range of 800 ppmv (compared to
current concentrations near 400 ppmv), rates of insect herbivory generally increase.
This is because the C:N ratio of plant tissues increases, and insects must consume
biomass at greater rates to obtain the nitrogen that often limits their growth and
fitness. There is also increasing evidence that increased temperature (especially
winter temperature) and more frequent droughts impose a stress on forests that
makes them more susceptible to insect herbivory and mass mortality. These stresses
increase the potential for insect larvae to overwinter in cold winter forests and thus
10
291
Fig. 5 Recent histories of insect outbreaks in temperate forest ecosystems of North America
(Redrawn from Hicke et al. (2012))
emerge with greater populations densities during the spring and summer, and they
compromise the capacity for trees to produce resinous (carbon based) and other
toxic (e.g., nitrogen based) herbivore defenses. Insect outbreaks in the temperate
forests of North America are episodic and have occurred several times during recent
decades (Fig. 5). One recent case of extreme disturbance in Western North American forests that has been linked to climate changes is that of the epidemic outbreak
of mountain pine beetles (Hicke et al. 2012).
The most recent mountain pine beetle epidemic originated in British Columbia,
Canada, at the beginning of the current century, and it has spread rapidly through
the Rocky Mountain region of the Western USA. It is important to note that this
insect (Dendroctonus ponderosae Hopkins) is native to montane pine forests in
Western North America and has emerged in small outbreaks in past decades. The
current outbreak, however, is especially large compared to previously recorded
events. The mountain pine beetle associates with fungal pathogens that are carried
by the beetles as they burrow into the bark of an infected pine tree and eat the
phloem tissue of the tree, creating infection galleries or tunnels that are
observable as the bark is peeled away. The fungal spores that are carried into the
trees vascular tissue by burrowing beetles often germinate and proliferate within
the phloem tissue, thus further disrupting the flow of sugars from shoots to the roots,
292
R.K. Monson
as well as penetrating into the xylem tissue and blocking water flow from roots to
shoots. Thus, an infected tree suffers from an inability to effectively transport
sugars to the roots to support respiration and growth and water to the shoot to
support transpiration. Trees typically die within 12 years after infection.
The current outbreak of mountain pine beetle in Pinus contorta (lodgepole pine)
forests in Western North America has caused over 14 million ha of forest to shift to
a state of mass mortality (Fig. 6). Infected stands typically exhibit over 6080 %
mortality of adult trees (Fig. 7). Two important secondary consequences have been
predicted to result from this outbreak: (1) large areas of the Northern Hemisphere
which had previously served as sinks for the uptake of atmospheric CO2 will serve
as sources of respired CO2 for numerous decades into the future as the dead needles
and wood decompose, and (2) forest fires will increase in frequency and coverage as
the dead needles and wood serve as fuel. However, studies of ecosystem processes
in beetle-damaged forests have revealed evidence that can be used to argue against
the likelihood of both of these long-term impacts. Recent research has shown that
widespread mortality of forest trees reduces the emission of CO2 from soil respiration to the atmosphere because the sugars normally transported belowground and
used to support both root and associated microbial respiration (often referred to as
autotrophic respiration) are reduced. Furthermore, even after needles are deposited
to the soil as litter, decomposition and the return of needle carbon to the atmosphere
appears to be limited, at least during the initial decade after forest death, by an
as-yet-to-be-identified resource or process (Moore et al. 2013). Thus, while carbon
budget models predict a large carbon source for beetle-killed forests, these model
predictions have not yet been validated by observations. In the case for increased
10
293
Fig. 7 (a) A lodgepole pine and aspen-dominated forest in the Rocky Mountains of the Western
USA showing the result of localized infection by mountain pine beetle. This forest is in the initial
stages of an epidemic outbreak. Aspen trees appear as the lighter shade of green. (b) Lodgepole
pine-dominated forest in the late stages of a mountain pine beetle infection in the Rocky
Mountains. This forest stand is in the transition between what is commonly referred to as the
red (earlier) and grey (later) stages of an infection. Note the presence of some live trees within the
stand. Mortality in these stands is typically between 60 % and 80 %
fire frequency in beetle-killed forests, most studies have not shown this prediction
to be borne out by observations, at least for the past two decades (Black et al. 2013).
It is probable that live pine trees, with their high needle and wood concentrations of
flammable terpene compounds, pose as great or greater fire risk to forests, than the
increased dead wood and needle deposition in forests with high mortality rates.
One disturbance that has been chronic, and anthropogenic, for at least the past
5 millenia, is human logging of temperate forests. Wood is a natural resource that
humans have used for a long time as both a fuel source and for the construction of
shelter. Deforestation in some parts of the world have increased over the past
several decades (e.g., the tropics), mostly for purposes of land use and the development of grazing-based animal husbandry and agriculture. However, in those
regions, such as Europe and North America, where temperate forests were once
used for high rates of log harvesting, removal of wood biomass has lagged behind
forest regrowth for at least the past five decades. This has allowed for increased
carbon sinks in the Northern Hemisphere. There is now concern that these sinks will
approach saturation and weaken. The exact causes of this weakening are numerous
and include climate changes and increased production of oxidant pollutants, such as
ozone, which tends to inhibit plant growth.
294
R.K. Monson
Future Directions
Determination of how temperate forest carbon sinks are responding to directional climate shifts, particularly with regard to decreasing snow packs in certain
regions at mid-latitudes, and increasing rain in other regions. The continentalscale redistribution of precipitation will undoubtedly impact carbon sequestration processes in temperate forest ecosystems.
Determination of how nitrogen deposition to forest ecosystems influences natural biogeochemical cycles involving nitrogen, carbon, and phosphorus.
Determination of interactions among disturbances on temperate forest biogeochemical cycling and the implications for such interactions on the future capacity for these ecosystems to provide water and take carbon out of the atmosphere,
two essential services provided to humanity by temperate forest ecosystems.
10
295
References
Black SH, Kulakowski D, Noon BR, DellaSala DA. Do bark beetle outbreaks increase wildfire
risks in the Central US Rocky Mountains? Implications from recent research. Nat Areas
J. 2013;33:5965.
Boon S. Snow ablation energy balance in a dead forest stand. Hydrol Process. 2009;23:260010.
Bugmann H. A review of forest gap models. Clim Change. 2001;51:259305.
Carcaillet C, Almquist H, Asnong H, Bradshaw RHW, Carrion JS, Gaillard MJ, Gajewski K, Haas
JN, Haberle SG, Hadorn P, Muller SD, Richard PJH, Richoz I, Rosch M, Goni MFS, von
Stedingk H, Stevenson AC, Talon B, Tardy C, Tinner W, Tryterud E, Wick L, Willis
KJ. Holocene biomass burning and global dynamics of the carbon cycle. Chemosphere.
2002;49:84563.
Chen IC, Hill K, Ohlem
uller R, Roy DB, Thomas CD. Rapid range shifts of species associated
with high levels of climate warming. Science. 2011;333:10246.
Clark DA, Brown S, Kicklighter DW, Chambers JQ, Thomlinson JR, Ni J, Holland EA. Net
primary production in tropical forests: an evaluation and synthesis of existing field data. Ecol
Appl. 2001;11:37184.
Davis MB. Lags in vegetation response to greenhouse warming. Clim Change. 1989;15:7582.
Hicke JA, Allen CD, Desai AR, Dietze MC, Hall RJ, Hogg EH, Kashian DM, Moore D, Raffa KF,
Sturrock RN, Vogelmann J. Effects of biotic disturbances on forest carbon cycling in the
United States and Canada. Glob Chang Biol. 2012;18:734.
Hogberg P, Read DJ. Towards a more plant physiological perspective on soil ecology. Trends Ecol
Evol. 2006;21:54854.
Hu J, Moore DJP, Burns SP, Monson RK. Longer growing seasons lead to less carbon sequestration by a subalpine forest. Glob Chang Biol. 2010;16:77183.
Kitzberger T, Brown PM, Heyerdahl EK, Swetnam TW, Veblen TT. Contingent Pacific-Atlantic
Ocean influence on multicentury wildfire synchrony over western North America. Proc Natl
Acad Sci USA. 2007;104:5438.
Lambers H, Raven JA, Shaver GR, Smith SE. Plant nutrient-acquisition strategies change with soil
age. Trends Ecol Evol. 2008;23:95103.
Lebreton P, Gallet C. Plant communities are biochemically organized. Acta Botanica Gallica
(in French). 2007;154:57395.
Lipson D, Nasholm T. The unexpected versatility of plants: organic nitrogen use and availability in
terrestrial ecosystems. Oecologia. 2001;128:30516.
Moore DJP, Trahan NA, Wilkes P, Quaife T, Stephens BB, Elder K, Desai AR, Negron J, Monson
RK. Persistent reduced ecosystem respiration after insect disturbance in high elevation forests.
Ecol Lett. 2013;16:7317.
Norby RJ, Wullschleger SD, Gunderson CA, Johnson DW, Ceulemans R. Tree responses to rising
CO2 in field experiments: implications for the future forest. Plant Cell Environ.
1999;22:683714.
Parmesan C, Yohe G. A globally coherent fingerprint of climate change impacts across natural
systems. Nature. 2003;421:3742.
Reich PB, Grigal DF, Aber JD, Gower ST. Nitrogen mineralization and productivity in tree stands
on diverse soils. Ecology. 1997;78:33547.
Saugier B, Roy J, Mooney HA. Terrestrial global productivity. San Diego: Academic; 2001. 573 p.
Schoennagel T, Turner MG, Romme WH. The influence of fire interval and serotiny on postfire
lodgepole pine density in Yellowstone National Park. Ecology. 2003;84:296778.
Thevenon F, Williamson D, Bard E, Anselmetti FS, Beaufort L, Cachier H. Combining charcoal
and elemental black carbon analysis in sedimentary archives: implications for past fire regimes,
the pyrogenic carbon cycle, and the human-climate interactions. Glob Planet Change.
2010;72:3819.
Vadeboncoeur MA. Meta-analysis of fertilization experiments indicates multiple limiting
nutrients in northeastern deciduous forests. Can J Forest Res. 2010;40:176680.
296
R.K. Monson
Van Tuyl S, Law BE, Turner DP, Gitelman AI. Variability in net primary production and carbon
storage in biomass across Oregon forests an assessment integrating data from forest inventories, intensive sites, and remote sensing. For Ecol Manage. 2005;209:27391.
Varhola A, Coops NC, Weiler M, Moore RD. Forest canopy effects on snow accumulation and
ablation: an integrative review of empirical results. J Hydrol. 2010;392:21933.
Vila M, Carrillo-Gavilan A, Vayreda J, Bugmann H, Fridman J, Grodzki W, Haase J, Kunstler G,
Schelhaas M, Trasobares A. Disentangling biodiversity and climatic determinants of wood
production. PLoS One. 2013;8, e53530.
Vitousek PM, Field CB. Ecosystem constraints to symbiotic nitrogen fixers: a simple model and its
implications. Biogeochemistry. 1999;46:179202.
Williams AP, Allen CD, Millar CI, Swetnam TW, Michaelsen J, Still CJ, Leavitt SW. Forest
responses to increasing aridity and warmth in the southwestern United States. Proc Natl Acad
Sci. 2010;107:2128994.
Further Reading
Box EL, Fujiwara K. Temperate forests around the Northern Hemisphere. Heidelberg: Springer;
2013. 280 p.
Frelich LE. Forest dynamics and disturbance regimes: studies from temperate evergreendeciduous forests. Cambridge: Cambridge University Press; 2008. 280 p.
Kurz WA, Dymond CC, Stinson G, Rampley GJ, Neilson ET, Carroll AL, Ebata T, Safranyik
L. Mountain pine beetle and forest carbon feedback to climate change. Nature.
2008;452:98790.
Turner MG, Romme WH, Gardner RH. Prefire heterogeneity, fire severity, and early postfire plant
reestablishment in subalpine forests of Yellowstone National Park, Wyoming. Int J Wildland
Fire. 1999;9:2136.
Wappler T, Labandeira CC, Rust J, Frankenhauser H, Wilde V. Testing for the effects and
consequences of mid-Paleogene climate change on insect herbivory. PLoS One. 2012;7,
e40744.
Plants in Deserts
11
Darren R. Sandquist
Contents
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Desert Formation Affects Desert Diversity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
The Abiotic Environment Underlying Desert Productivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Precipitation and Drought . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Functional Diversity and Responses to the Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Ecological Groupings of Desert Plants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Photosynthesis in a Water-Limited Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Adaptive Forms and Functions Related to Desert-Plant Water Relations . . . . . . . . . . . . . . . . .
Biotic-Mediated Processes Are Critical for Nutrient Balance in Deserts Plants . . . . . . . . . .
Desert Biodiversity and Community Composition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Species Diversity Can Be Surprisingly High in Deserts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Population and Community Dynamics Are More Complex than Expected . . . . . . . . . . . . . . .
Disturbance, Global Changes, and Future Challenges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Disturbances Pose Significant Challenges in Low Productivity Ecosystems . . . . . . . . . . . . .
Nonnative Species Are a Major Threat to Desert Communities . . . . . . . . . . . . . . . . . . . . . . . . . . .
Other Global Changes also Threaten Desert Regions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
299
300
301
302
302
302
303
311
317
319
319
321
324
324
324
325
325
Abstract
There is no single definition of desert, but it is widely agreed that deserts are
arid because they receive little precipitation and experience high evaporation
annually. These factors result in low soil water availability that severely limits
plant productivity. Thus, another feature of deserts is low vegetation cover.
Although all deserts are dry, there is extreme abiotic and biotic variability
among the worlds deserts perhaps more so than for any other biome. This
arises in part from the varied causes of desert formation, their disjunct
distributions, and their independent floral histories.
D.R. Sandquist (*)
Department of Biological Science, California State University, Fullerton, CA, USA
e-mail: dsandquist@fullerton.edu
# Springer Science+Business Media New York 2014
R.K. Monson (ed.), Ecology and the Environment, The Plant Sciences 8,
DOI 10.1007/978-1-4614-7501-9_3
297
298
D.R. Sandquist
11
Plants in Deserts
299
Introduction
Desert is the biome classification for terrestrial regions of Earth that are climatically
arid and have low vegetation cover. Additionally, the climate of such regions is
often highly variable across seasons and years. While there is no single index that is
used universally to define deserts, a simple one, proposed by Miegs (1953), is based
only on precipitation, whereby extremely arid regions experience at least 12 months
without rainfall, arid regions receive <250 mm rainfall annually, and semiarid
regions receive 250500 mm rainfall annually (Fig. 1). Boundaries based on this
index do a good job delimiting deserts across the globe and correspond closely to
boundaries used in other classification systems (e.g., Ezcurra 2006). But aridity is
not simply based on the amount of water derived from precipitation; it also depends
on the loss of that water, which affects its availability for plant productivity. A more
inclusive definition of aridity comes from the comparison of water loss via evapotranspiration (ET) versus water input from precipitation (P). The ratio P/ET is a
commonly used index of aridity (e.g., UNESCO 1977) defining hyperarid zones as
having P/ET <0.03 and arid zones having P/ET of 0.030.20. Although this
definition does not significantly change the global boundaries of deserts as compared to other indices, such as Meigs, it does provide a more biologically relevant
measure of aridity in terms of water availability for plant use.
Other environmental parameters, such as timing and intensity of rainfall, seasonal temperatures, and soil texture, to name a few, can also play a role in affecting
the aridity of deserts, albeit at smaller spatial and temporal scales. These additional
factors affect the abiotic heterogeneity within deserts that contributes to the surprising functional diversity of plants found in desert ecosystems. This chapter explores
the diversity of desert plants from an ecological context. It begins with a short
review of desert formation and abiotic variability as a foundation for understanding
the causes of biotic diversity among and within deserts. Then, the diversity of desert
vegetation is explored from a functional context through the community level.
Sonoran
Lut
30
Taklimakan
Sahara
Arabian
Thar
Chihuahuan
Somali-Chatbl
Atacama
Namib
Hyperarid
Arid
Semiarid
0
2000 km
30
Kalahari
Great Sandy
Simpson
Patagonian
Fig. 1 Global distribution of nonpolar arid lands based on Meigs (1953) classifications
300
D.R. Sandquist
It ends with some considerations of how the desert biome has changed due to human
activities and how it may change with future global changes.
11
Plants in Deserts
301
opposite side, the air pressure and temperature rise (similar to Hadley cells). The
limited moisture left in the descending air mass is prevented from precipitating
because of the increasing pressure and temperature. Low rainfall, warm air, and
high solar radiation resulting from the presence of a mountain range result in arid
areas in the mountains shadow.
Coastal deserts form where very cold ocean waters occur at the surface and
adjacent to a relatively warm continental margin (e.g., the Namib Desert occurs
where upwelling brings cold water to the surface along the western coast of Africa).
The interaction of the ocean, air, and land is complex in these systems, but in
general the cold surface waters cause air masses that overlie them to cool. This
decreases evaporation and reduces the capacity of the overlying air to hold water
vapor, causing condensation and offshore precipitation. Sometimes the condensation forms fog, which may be drawn onto land, but as the fog blankets the land, it
too warms and evaporates back into vapor. Because of this phenomenon, coastal
deserts may also be known as fog deserts. Coastal deserts are among the driest in the
world sometimes experiencing years without measurable rainfall (e.g., Namib and
Atacama). In fact, fog is typically the most reliable source of water for productivity
in these deserts, and many plants of these deserts show adaptations for capturing
and taking up fog-derived water (e.g., Nolana mollis).
Besides the different formation processes, highly varied ages among the worlds
deserts also contribute to the diversity among them. Some deserts appear to be
extremely old (e.g., Namib Desert >55 myo) giving rise to high diversity and
endemism through many generations of evolution. Other deserts are very young
(e.g., Mojave Desert ~11,000) and are strongly impacted by migration processes
from regional biota. This can also lead to high diversity, especially if the desert
forms at the intersection of multiple ecological regions. Biodiversity among deserts
is also a result of their disjunct distribution. That is, deserts of the world are largely
separated from each other compared to other biomes. As such, evolutionary processes within them have taken place largely in isolation from each other.
302
D.R. Sandquist
11
Plants in Deserts
303
water stress. Drought-deciduous shrubs (plants that do not die, but do lose their
leaves during drought) are drought enduring. There is overlap in Shantz (1927)
definitions of drought evading and drought resisting, which appears to have generated some confusion in the literature. Both types are active during drought, but
drought-evading plants typically have higher growth per unit water used (i.e.,
higher water-use efficiency) due to adaptive traits that reduce water loss and
prolong the growing period. Reduced transpiration due to stomatal regulation
coupled with morphological features such as stomatal pits, leaf hairs or waxes,
and small leaf sizes is often found in drought evaders. Kearney and Shantz (1912)
also classified plants with extensive root systems into the drought-evading category.
Drought-resisting plants have persistently low-to-moderate levels of activity
through periods of low water availability as well as during more favorable periods.
Reduced transpiration is the norm for drought resisters, but they can tolerate very
low water potentials, often via osmotic regulation. Succulent plants and some of the
most successful desert perennial shrubs (e.g., creosote bush) fall into this category.
Categorizing plants in terms of their functional attribute is useful only to a limited
extent, and there are many examples of taxa that exhibit properties of more than one
category. For example, one could argue that creosote bush exhibits both droughtresisting and drought-enduring characteristics (small leaves with resinous excretions
to reduce transpiration). For this reason, a popular alternative is to group desert plants
based on life forms. These forms usually include annuals, perennial grasses, deciduous shrubs, evergreen shrubs, CAM succulents, and deep-rooted trees (phreatophytes). Smith et al. (1997) took this approach in their summary of North American
desert plant ecophysiology, and many others have applied it as a way to simplify the
presentation of the complex diversity of form and function found in desert species.
Interestingly, these diverse life forms are present in a broad cross section of desert
taxa suggesting that the mechanisms for dealing with aridity and heat, or the ability to
form them through natural selection, are fundamental to many lineages.
304
D.R. Sandquist
foundation for many of the adaptive characteristics seen in desert (and other) plants.
In deserts, plant water loss from leaves is exacerbated by high water vapor pressure
differences (VPD) between the leaf and the air. When water is abundant, as during the
growing period for annuals, the water lost via transpiration may be inconsequential,
especially compared to its role in reducing leaf temperatures. For plants that remain
active during the drought period, mechanisms that help reduce water loss should be
favored. The most straightforward mechanism for reduction of water loss during
periods of drought is to reduce the size of the stomatal opening, thereby decreasing
conductance of water from the leaf. But this comes at a cost; it reduces uptake of CO2.
In addition, for most desert plants, a reduction in transpiration results in a potentially
dangerous increase of leaf temperature (see discussion of energy balance in chapter
Plants in Alpine Environments). Over the years, myriad fascinating examples of
morphological, physiological, and behavioral mechanisms have been identified that
help desert plants avoid the full consequences of these trade-offs. In general these can
be grouped into ways of improving photosynthesis relative to water loss, decrease
dependence on transpiration for energy balance, and ways to take up or save more
water.
Photosynthetic Pathways
Pick up almost any book about photosynthesis and entire chapters can be found
about C3, C4, and CAM photosynthesis. Indeed, the ecology and biochemistry of
these three photosynthetic pathways differ so greatly that they warrant entire
volumes. Rather than review the three photosynthetic pathways in detail, the
attributes of each that are important for their presence in deserts are highlighted;
then their distribution and how the different pathways correspond to variability
among these arid ecosystems are explored. All three pathways are present across
the deserts of Earth, but as might be expected, their abundances differ in relationship to the environments of each desert.
Of the three pathways, C3 photosynthesis is the most widespread globally, and
the same is true across deserts. However, net carbon gain of C3 plants is negatively
affected by photorespiration, which goes up with increasing temperatures. This is
one reason C4 and CAM plants may have a competitive advantage over C3 plants in
hot deserts (Ehleringer and Monson 1993). In deserts with cooler temperatures
during the growing season, the disadvantage of photorespiration is significantly
lower, thereby reducing the relative benefit of the C4 pathway. The C4 pathway
also requires two additional ATP to fix CO2 (compared to C3), making it best suited
to high-light environments. As expected from these fundamental differences, the
greatest abundance of C4 plants in deserts is where temperatures and light are high
and water is available during warm periods.
C4 plants have high water-use efficiency (carbon gain vs. water loss) because the
CO2-concentrating mechanism of the C4 pathway maintains higher internal CO2
concentrations relative to stomatal conductance and thus transpiration. However,
the need for water during the warm growing season prevents most C4 plants from
being well adapted to drought conditions. In contrast, CAM plants have extremely
high water-use efficiency. They benefit from the same CO2-concentrating
11
Plants in Deserts
305
mechanism found in C4 plants, but additionally keep their stomata closed during
the day when evaporative demand (i.e., VPD) is high, and then open them for CO2
exchange during the lower-VPD hours of darkness.
Other attributes of CAM species benefit their tolerance of drought. As previously
described, many CAM plants are succulent, having tissues that store water for
use during drought, but CAM is not restricted to succulent plants. Likewise, not
all succulents are CAM (e.g., many leaf-succulent shrubs of the Succulent Karoo
are C3). Some CAM species can switch to C3 photosynthesis when environmental
conditions are favorable, especially when water availability is high, and many
can use C3 photosynthesis during a small fraction of the regular daily CAM cycle.
CAM plants increase in abundance in hot deserts that have some degree of water
limitation during the warm season. This limitation may stem from an absence of
precipitation or from an ephemeral and unpredictable precipitation regime. But,
most CAM species are sensitive to freezing temperatures and thus absent from cold
(high-elevation or high-latitude) deserts.
In North American deserts, the relative abundances of C3, C4, and CAM along a
north-to-south gradient of increasing temperature and summer rainfall reflect the
typical pattern among global deserts. In the winter-rain-dominated Great Basin cold
desert, CAM and C4 plants are largely absent except in saline habitats (see Desert
Halophytes). The proportion of CAM and C4 species increases slightly in the
Mojave Desert to the south, where annual temperatures are warmer but winter rains
still dominate. In both of these deserts, C3 species greatly outnumber C4 and CAM
species. Even further south, and at overall lower elevations, CAM species become
an important part of the Sonoran Desert flora, with some taxa (e.g., Cactaceae)
showing remarkable morphological as well as taxonomic diversity. Summer rains
are abundant here but spatially and temporally variable. C4 species, especially
grasses, also become a more integral part of the flora in the Sonoran Desert but
normally in the higher elevations where rainfall is more abundant and predictable.
The southernmost North American desert, the Chihuahuan, has an abundance of
CAM and C4 species related to the higher annual temperatures and summer rainfall
of this desert. CAM agaves and cacti are more speciose here and can be the
dominant taxa of some Chihuahuan communities. C4 grasses can likewise dominate vast areas of the Chihuahuan, especially where rainfall is relatively plentiful.
But, both C4 grasses and CAM species are often not the dominant plants on heavily
calcareous soils that occupy many parts of the Chihuahuan. Here they are replaced
by C3 shrubs (primarily creosote bush and tarbush) a shift that probably reflects
poor retention of shallow water on such substrates. This pattern illustrates the
potential for local edaphic effects to modify climate patterns that would otherwise
favor certain ecophysiologcial syndromes over others.
306
D.R. Sandquist
latent heat transfer poses few, if any, problems. But when water is limited, which
often corresponds to the warmer periods of the year, reliance on latent heat transfer
presents a challenge. This challenge appears to have driven functional diversification and adaptation at the leaf level among many desert plants, as well as in other
ecosystems. (For a more detailed review of energy balance, see the chapter
Plants in Alpine Environments.)
Small Leaves Decrease Leaf Temperature and Transpiration
Reduced leaf size is one of the most widespread morphological adaptive features
seen in desert plants. It seems intuitive that because there is less surface area on
smaller leaves, water loss will be lower, but this is not necessarily true. Water loss
from a leaf is dependent on transpiration rate, which is an area-standardized measurement (e.g., mmol H2O m 2 leaf s 1). A priori, small and large leaves can have
the same transpiration rate, in which case a canopy of many small leaves will lose the
same amount of water as one with fewer large leaves (i.e., the total surface area is the
same). For small leaves to be adaptive in terms of water loss, they must instead have
a lower transpiration rate, which, as explained below, they usually do. Small leaves
also do not heat up to the same extent as larger leaves. These two properties go handin-hand, and since heat and water limitations are two of the greatest challenges for
desert life, it is not surprising that small leaves are common in the desert flora.
The primary reason smaller leaves stay cooler, and subsequently have lower
transpiration than larger leaves, is that they have a reduced boundary layer for heat
transfer. A smaller boundary layer means that heat transfer from the leaf to the
surrounding air (i.e., convective heat transfer) is more rapid. Thus, as the leaf heats
up from absorption of radiation from the sun and surrounding objects, higher
convective heat loss keeps the leaf temperature closer to the air temperature
(T). Convective heat loss means that the plant is less dependent on latent heat
transfer, via transpiration, for maintaining a favorable leaf temperature. But additionally, a lower T also reduces the vapor pressure difference (VPD) between the
leaf and air, which also lowers transpiration.
Lower leaf temperature may also benefit the leaf in terms of photosynthetic rate
since the lower temperature is likely closer to the thermal optimum for photosynthesis. Recall also that lower temperatures reduce photorespiration in C3 plants.
For many species, leaf sizes can vary across seasons and years, with smaller
leaves produced during warmer periods or during drought. Such adjustment
are crucial in plants that persist through periods of water shortage and high
temperatures, underscoring the importance of another adaptive function in desert
plants acclimation.
Leaf Angles and Leaf Movement Affect Light Interception
Another beautiful example of acclimation in desert plants is leaf movement known
as heliotropism (meaning sun orienting). Some desert species, mainly annuals,
display diaheliotropic leaf movement (orientation perpendicular to sun rays) and
paraheliotropic leaf movement (orientation parallel to sun rays), although not all
species do both. The former maximizes interception of solar radiation whereas in
11
Plants in Deserts
307
Fig. 2 Interception of solar radiation (measured as photon flux density) by diaheliotropic and
paraheliotropic leaves during daylight hours. Arizona lupine (Lupinus arizonicus) of the Mojave
and Sonoran Deserts can switch from fully diaheliotropic during periods of favorable soil moisture
to fully paraheliotropic during water-stressed periods or combine both dia- and paraheliotropism
during a single day. For comparison, interception by a non-heliotropic horizontal leaf is also
shown.A vertical leaf (not shown) would have an inverted curve from the horizontal leaf (Redrawn
with permission from J. R. Ehleringer)
308
D.R. Sandquist
Fig. 3 Copiapoa cinerea ssp. columna-alba of the Atacama Desert, Chile, grow with a northward
orientation that helps maintain warm temperatures on the apical meristem during the cool period of
the year but reduces heat load during the hot season (Photo: D. R. Sandquist)
angles function much the same as switching between dia- and paraheliotropism
(Fig. 2). That is, they maximize light capture in early morning and late afternoon
hours, when air temperature and VPD are lower, but avoid direct solar radiation in
the more severe midday hours. Nonrandom leaf orientations may also reduce selfshading, with the angles being specific not just to daily radiation changes but also to
seasonal changes. Such orientation benefits are also found in photosynthetic stems,
including those of succulent species. An example of this is seen in the cactus
Copiapoa cinerea ssp. columna-alba from the Atacama Desert of Chile. The succulent stems of these plants orient due-north giving the comical appearance of a small
cactus army marching towards the equator (Fig. 3). Ehleringer et al. (1980) showed
that this orientation facilitates apical warming for growth during the cool/wet parts of
the year and reduces radiation (thus heat load) during the driest part of the year.
Reflective Leaf Surfaces Decrease Absorption of Solar Radiation
The multiple benefits of reducing direct solar radiation suggest that other leaf
properties should serve this function in desert plants. Indeed, there are a number
of traits that do so at the leaf surface, reflective waxes and leaf hairs being among
the most common. A well-studied example of this is found in brittlebush (Encelia
farinosa, Asteraceae), a drought-deciduous shrub of the Mojave and Sonoran
Deserts. Leaves produced by this species can have a thick layer of trichomes (leaf
hairs) that strongly reflects solar radiation. Notably, the thickness of the trichomes,
and thus the amount of reflectance, depends on the level of water stress experienced
by the plant. Leaves produced early in the rainy season are generally large and have
11
Plants in Deserts
309
Fig. 4 Micrographs of
brittlebush (Encelia farinosa)
leaves from the Mojave
Desert. (a) Leaves produced
early in the growing season
when soil water availability is
favorable have low trichome
densities. (b) Leaves
produced later in the season,
when water stress has
increased, have a dense
trichome layer (Photos: J. R.
Ehleringer)
few trichomes (Fig. 4a). These leaves absorb ~80 % of the solar radiation incident
on their surface, but the heat load resulting from this radiation is easily balanced by
transpiration during this wet period of the year. As the season progresses, and soil
water decreases, new cohorts of leaves are produced which have increasing trichome densities (Fig. 4b). The higher densities lower radiation absorption to as
little as 40 %, which attenuates excessive heat load and, importantly, reduces
dependence on transpiration as the plants enter the drought period. As one would
expect, the lower light absorption also decreases photosynthesis, but acclimation
through increased trichome development allows plants to remain active much
longer into drought, thereby compensating for the decrease of photosynthesis.
Biochemical Acclimation Changes Thermal Optimum of Photosynthesis
Rather than maintaining a narrow range of leaf temperatures for optimal photosynthesis, an alternative is to change the optimum temperature. (One might call this,
310
D.R. Sandquist
if you cant beat them, join them.) In a number of desert species, biochemical
adjustments do just this. Such physiological acclimation results in changes of the
optimum temperature for photosynthesis that closely match seasonal differences in
ambient temperatures (Fig. 5). Thermal acclimationof photosynthesis is found
primarily in evergreen plants across many forms (e.g., shrubs, grasses, succulents,
and ferns) and appears to be uncommon in annuals and drought-deciduous perennials, presumably because these two growth forms do not experience the breadth of
leaf temperatures that evergreen species do.
Creosote bush (Larrea tridentata, Zygophyllaceae) is often cited as the quintessential thermal acclimating desert species, showing temperature optima changes in
the field from 20 C in January to 32 C in September. Importantly, these changes
could be replicated in reciprocal transplant and controlled temperature experiments,
thereby confirming the response to be acclimation based specifically on
temperature.
11
Plants in Deserts
311
When water is abundant, the majority of plants with photosynthetic stems flush a
cohort of small leaves that have high transpiration and photosynthetic rates. As
drought ensues leaves are abscised but carbon gain continues in the photosynthetic
stems. These stems usually have lower net photosynthetic rates but greater wateruse efficiency than leaves. Some studies have also shown stems to have higher
temperature optima for photosynthesis than leaves. These trends suggest that these
highly modified photosynthetic stems are well adapted for operation during dry and
potentially hot conditions of deserts, enabling year-round carbon gain for many
species and potentially facilitating more rapid responses to pulses of water
availability.
The costs associated with stem photosynthesis (e.g., construction costs and lower
carbon assimilation) may be high compared to photosynthetic leaves, but those
costs appear to be outweighed by the benefits. For some species the contribution of
photosynthetic stems and twigs to annual plant carbon gain is important, as it can
exceed 70 % (Szarek and Woodhouse 1978) and extend carbon uptake by 7 months
(Tinoco-Ojanguren 2008). Furthermore, stems play other structural roles that
should also be considered in the benefits, as not all plants with green photosynthetic
stems engage in exogenous gas exchange. Instead, these species benefit from stem
photosynthesis through the re-fixation of respired CO2, which may help maintain
reserves of stored carbohydrates.
312
D.R. Sandquist
Day
night
day
90
Air
30
4.2
Leaves
1.2
4.0
Surface
soil
4.0
4.1
Roots
1.2
1.0
Deep
soil
1.0
Night
Hydraulic lift
Transpiration
Fig. 6 Hydraulic lift is the process of water being moved from areas in the soil with high water
potential to areas with low water potential via plant roots. This occurs at nighttime when stomata
are closed and transpiration of water from the leaves is shut down. The translocated soil water may
serve as a reserve for plant uptake the next day when transpiration resumes
Such high productivity is partly due to phreatophytes being largely decoupled from
surface drought conditions. Such decoupling allows phreatophytes to be productive
throughout rainless periods, even when those rainless periods extend for years as
for Prosopis tamarugo, a deeply rooted species of the hyper-arid Atacama Desert of
Northern Chile.
Deep roots are also commonly found in plants that display a functional process
described as hydraulic lift (or hydraulic redistribution). Popular accounts of this
process describe it as self-watering by plants, whereby water from zones of high
water potential (usually deep soil) is nocturnally redistributed through roots to
zones of low water potential (i.e., shallow soils) and stored there until daytime
when the plant takes up the stored shallow water for transpiration. This phenomenon was first quantified in the ubiquitous Great Basin Desert shrub Artemisia
tridentata (big sagebrush) and coined hydraulic lift because water movement
was in an upward direction. It has since been found in other desert species with
roots that experience a hydraulically heterogeneous soil profile. In spite of the
apparent fitness value of hydraulic lift, its presence has not been widely examined in
most ecosystems, including deserts.
Hydraulic lift (Fig. 6) is driven by water potential differences that develop in the
soil profile during the day. Root densities are typically greatest in shallow soil and
decrease with depth, as such transpiration depletes water in the shallow layers to a
greater extent than that at depth (especially if the deep roots are near saturated
soils). Evaporation also contributes to a higher loss of water from shallow soils. The
result is a water potential gradient in the soil profile that is bridged by the roots of
the plant. At night, when stomata close and transpiration is greatly reduced, water
continues to move within the plant along the residual water potential gradient.
11
Plants in Deserts
313
Movement continues from the roots to the shoots until the shoot water potential
comes to equilibrium with the root water potential. Below ground, in a similar
manner, water fluxes from the deep roots in moist soil (high water potential) to the
shallow roots in soil that has dried through the day (low water potential). The
surprising aspect of this process is that the water leaks out of the shallow roots
and into the surrounding dry soil, meaning that water movement in these shallow
roots reverses. The hydraulically lifted water accumulates in the shallow root zone
overnight, where it is then available for uptake by the plant the next day when
transpiration resumes and the shallow root flux again reverses.
As might be expected, water redistribution by roots can also occur in the
opposite direction inverse hydraulic lift. The movement of water from wet
upper soil layers (such as after a monsoon rain) into dryer deep layers may benefit
root growth into deep soil and can redistribute water away from access by shallowrooted competitors. Inverse hydraulic lift has been demonstrated for a number of
desert plants and across life forms, including Kalahari dune grasses and a
Chihuahuan tree (Arizona walnut). The facultative phreatophyte, Prosopis velutina
(velvet mesquite) of the Sonoran Desert has even been shown to engage in both
hydraulic lift and inverse hydraulic lift.
Recently, hydraulic lift was shown to facilitate greater nutrient availability to
plants. This results from microbial activity in shallow root zones being stimulated
by hydraulically lifted water exuded by the plant. This important discovery adds
another dimension (self-fertilization) to the value of hydraulic lift.
314
D.R. Sandquist
Fig. 7 Leaf and stem succulent species dominate the vegetation of the Vizcano region of the
Sonoran Desert. Centered in this photo is the elephant-stemmed Pachycormus discolor. The large
columnar cactus to the left is Cardon, Pachycereus pringlei, and the tall slender plant to the right is
Boojum tree (Fouquieria columnaris). A number of other succulents from the Agavaceae and
Cactaceae families are also present in this scene (Photo: D. R. Sandquist)
11
Plants in Deserts
315
from unfavorable environmental conditions (e.g., low soil water, high soil salinity)
and facilitates growth and survival through such periods. Cacti are arguably the
most recognizable of the succulent species with their fattened stems and spiny
armor, but succulence is present in 30 of 50 plant orders (Eggli and Nyffeler 2009),
most of which are represented in arid and semiarid habitats. These taxa show great
diversity of form and physiology in spite of having the similar ultimate function of
attenuating water stress.
In places, succulents may dominate the biomass or diversity of a desert. The arid
and semiarid regions of southern Africa that include the Namib Desert and Succulent Karoo harbor the highest diversity of succulents equaling approximately 1/3
of the estimated ~10,000 succulent species globally. Parts of the Sonoran Desert are
so influenced by the presence of succulent species that Forrest Shreve relied heavily
on them for delineating four of six vegetational subdivisions (Shreve and Wiggins
1964). The Arizona Upland is crassicaulescent (succulent stem cacti), the Central
Gulf Coast is sarcocaulescent (fleshy stem trees), the Vizcano region is
sarcophyllous (succulent leaf), and the Magdalena region is arbocrassicaulescent
(tree and stem succulent). These divisions also highlight the most common groupings of succulence among species: leaf succulents (e.g., aloe and yucca), stem
succulents (e.g., most cacti), and caudiciform succulents, whose succulent parts
may included non-photosynthetic portions of the stem, the upper part of the root and
the root proper (e.g., many Euphorbia sp.).
Given the pulsed nature of rainfall in most arid ecosystems, rapid uptake of large
quantities of water is important for succulent species. To accomplish this, many
succulents have extensive root systems that are often only a few centimeters below
the soil surface (e.g., cholla and barrel cacti). Another adaptive feature of the roots
of some succulent species is the very rapid formation of new roots when water is
present. These rain roots form within a couple days of wetting and die once the soil
is again dry. Thereafter the main root system is impermeable to water uptake and
water loss throughout the dry period, which can last for many months. Both shallow
roots and rain roots provide a mechanism for succulent plants to rapidly take
advantage of ephemeral water availability and small rain events that wet just the
upper soil layer.
In leaf and stem succulent plants, water is typically stored in the vacuoles. Thus,
another feature of succulent plants is the presence of very large vacuoles in the
succulent tissues, occupying up to 90 % of the cell volume. These vacuoles also
serve another purpose in many succulent species, storage of organic acids associated with CAM photosynthesis. Most succulent species display some degree of
CAM photosynthesis (although there are many without any CAM activity). The
combination of CAM and succulence represents a structure-function relationship
that is remarkably well suited for life in warm and arid environments.
316
D.R. Sandquist
high evaporation rates that exceed precipitation input. Dissolved solutes are not
leached from these soils; instead, they are concentrated near the soil surface as
water evaporates. In low-lying basins, salts are transported with rain runoff from the
surrounding elevations and then further concentrated by evaporation. Over many
years, this process results in extremely saline playas or salt basins. The center of
most basins has such high salinity and fine soil particles that no vegetation can
establish or survive, but along the margins where particle sizes are larger and
salinity is not as extreme, the plant community is usually unique, composed of
species that can survive relatively high salinity. Such salt-adapted plants are known
as halophytes, meaning salt plants.
Plants that live in saline habitats but have mechanisms to prevent uptake of salts
through the roots are called salt avoiders or excluders. These are not true halophytes
because they always grow best in the absence of salinity. In general, salt excluders
are not particularly common in deserts because the process of salt exclusion leads to
increasingly greater soil salinity in the rooting zone.
True halophytes take up salt minerals (primarily Na+, K+, and Cl ) through the
roots and into the plant tissues; thus they face the challenge of preventing physiological dysfunction and possible cell death caused by the toxicity of high salt
concentrations. Controlled balance of cell ionic concentrations through rapid growth
and synthesis of compatible organic solutes (i.e., osmotic adjustment), coupled with
compartmentalization of the salt ions are keys to salt tolerance in halophytes.
Another challenge to growth in saline soils is that salinity causes soil water potential
to be lower, making it more difficult for plants to take up water. For halophytes,
however, the uptake of salts into the roots facilitates water uptake by lowering the
root water potential, thereby counteracting the problem of lower soil water potential.
Some halophytes actually have lower growth in nonsaline soils than in those
with modest levels of salinity (i.e, 50250 mM NaCl) (Flowers and Colmer 2008).
All, however, must have mechanisms to prevent the toxic ramifications of high salt
concentrations in living tissues. Salt accumulators prevent these negative effects by
sequestering salts in the vacuoles or other cell structures, thus eliminating interactions between the salts and cytoplasmic components and membranes. Many salt
accumulators are succulent because they rely on large vacuoles for this purpose.
Examples of succulent salt accumulators are common in the Chenopodiaceae
family (e.g., Salicornia, Suaeda, and Allenrolfea), but also from this family are
species in the genus Atriplex that sequester salts in modified epidermal hairs (salt
bladders). Interestingly, the salt bladder can serve an additional beneficial function.
As water evaporates from the bladder, the salts precipitate from solution and
become white. This increases the albedo of the leaf, which, like the leaf hairs of
Encelia farinosa, increases leaf reflectance of solar radiation, attenuates heat load,
and reduces transpiration (Mooney et al. 1977).
Another mechanism to avoid the toxic effects of high salinity is to excrete the
cellular salts onto the outer leaf or stem surface. Salt excretors are found across plant
functional groups and taxa (e.g., salt cedar tree, Tamarix, and salt grass, Distichlis).
Many rely on specialized salt glands to excrete the cellular salts, where once on the
surface, the salt is either washed or blown off or eliminated when the leaf abscises.
11
Plants in Deserts
317
318
D.R. Sandquist
Among the most widespread are trees in the legume family (Fabaceae), which form
nitrogen-fixing associations with Rhizobium and Bradyrhizobium bacteria. These
trees, and the input of nitrogen due to their presence, are important components of
desert communities across both the western and eastern hemispheres (e.g., Acacia in
African and Middle East deserts; Prosopis in North and South American deserts).
Nitrogen-fixing associations may also be formed between actinomycetes and plants
and between free-living bacteria and plants that release root secretions into the
rhizosphere surrounding roots. The latter is often accompanied by a rhizosheath, an
anatomical specialization of the root that facilitates development of the bacterial
association. The importance of these alternative nitrogen-fixing associations in
deserts is poorly understood at present.
Biological soil crusts play many important roles in arid and semiarid ecosystems, including nitrogen input through nitrogen fixation. Biological soil crusts are
an autotrophic microbiotic community composed of cyanobacteria (and other
bacteria), green algae, lichens, mosses, and microfungi. Organisms in these
microcommunities grow together as a mat or mound that integrates with particles
in the top few millimeters of the soil via a network of cyanobacteria and fungal
filaments. All arid and semiarid regions of the world have biological soil crusts, and
in some places they occupy up to 70 % of the surface cover. In places where such
crusts are present, plants often have greater overall biomass and higher tissue
nitrogen concentrations (e.g., tissue nitrogen is 931 % higher for plants growing
among biological soil crust in the Great Basin Desert).
One function of biological soil crusts that contributes to plant nutrition is fixation
of atmospheric nitrogen by cyanobacteria and lichens of these microcommunities.
This nitrogen is made available to plants through both decomposition of dead
biomass and leaking of nitrogen from the cyanobacteria and lichen. For example,
the cyanobacteria Nostoc has been shown to lose up to 80 % of its fixed nitrogen.
This nitrogen enters the soil mostly as NO3 and is readily available for plant
uptake. However, fixation and release of nitrogen are highly variable within and
between deserts depending on the species composition of the crust, the soil moisture levels, and the soil temperatures. Biological soil crusts also contribute carbon
to the soil microbial communities of deserts, thereby benefiting decomposition and
other microbial-mediated processes that impact plant nutrition.
Biological soil crusts may also affect desert plant communities because of their
impact on soil water availability, seed germination, and plant establishment. Soil
water is typically greater in the presence of biotic crust because it slows the surface
movement of water, which allows greater time for infiltration and may reduce
evaporation from the subsurface. These benefits are best realized after large or
prolonged precipitation events. Small pulses of rain may only wet the biotic crust
without ever percolating into the subcrust soil. A number of studies have also
shown biological soil crusts to improve or, at worst, not affect germination and
establishment of native plants. In contrast, many alien species have reduced germination and establishment on biotic crusts. Such findings imply an evolutionary
response of native desert plants to the presence of biological soil crust, but few
hypotheses based on this context have been tested.
11
Plants in Deserts
319
Mycorrhizal fungi facilitate uptake of water and nutrients of desert plants in the
same manner as they do for other species, and they appear to be as widespread
among desert plant families as for those of other ecosystems. Most desert mycorrhizae are of the arbuscular type. They improve uptake of water and nutrients
because the extensive network of fungal hyphae greatly increases the functional
surface area for uptake while exploiting a greater soil volume than do roots alone.
Dark septate endophytic fungi is another group of fungi that form associations with
desert plants. This group appears to be equally wide spread in deserts as mycorrhizae, perhaps more so. Their prevalence has led most authors to conclude that they
play an important role in desert ecosystems and for desert plants (including for
water or carbohydrate storage) yet their exact role has been difficult to elucidate.
320
D.R. Sandquist
Fig. 8 The Succulent Karoo is considered a global biodiversity hot spot. It harbors over 5,000 plant
species including about one-third of the worlds succulent species (Photo courtesy of A. G. West)
to its high biodiversity. Winter low temperatures are not extreme and neither are
summer high temperatures; the latter are often buffered by cool coastal fogs and
dew, which also serve to reduce drought severity and duration. Rainfall, though
low (150 mm on average), comes in winter, and unlike many other deserts, it is
relatively predictable.
11
Plants in Deserts
321
The central Sahara massifs, for example, receive enough water to form shrubland
and grassland communities that are different from the surrounding desert and often
contain endemic taxa and where water persists for long periods of time, unique
montane wadis communities form.
322
D.R. Sandquist
More direct evidence of competition has been elucidated from field manipulations of plants and plant resources. One early and widely cited example includes
Larrea tridentata (creosote bush) and Ambrosia dumosa (white bursage), two of the
most ubiquitous species of North American deserts. Widespread coexistence of
these two species led to questions about belowground resource competition. In a
series of studies, Fonteyn and Mahall (1978, 1981) used different combinations of
plant removals (e.g., removal of Larrea only, Ambrosia only, neither, or both) to
demonstrate both the presence of interspecific competition for water and that
intraspecific competition was weaker than interspecific competition. Later studies
by this group identified two mechanisms for these outcomes: inhibition of root
growth mediated by an apparent exudate from Larrea roots (allelopathy) and
avoidance of overlapping growth due to physical contact between roots of Ambrosia. Since these studies, a number of other neighbor-removal experiments in deserts
have confirmed that competition, especially for water, is common both within and
among species and for plants showing regular, clumped, and even random
distributions.
Interspecific competition is one mechanism expected to lead to resource
(or niche) partitioning between species. Studies on coexistence among desert plants
have been instrumental in testing and, in many cases, verifying this concept, and
although resource partitioning may not lead to full elimination of competition, it
helps to minimize it. One widely used framework for such investigations is the
Walter two-layer hypothesis attributed to German ecologist Heinrich Walter
(1939, reviewed in Ward et al. 2013). The two-layer hypothesis predicts that
species may coexist by partitioning belowground water resources such that one
species relies primarily on shallow soil water and the other on deeper soil water.
Originally proposed to explain coexistence of savanna grasses (shallow rooted) and
trees (deeper rooted), this model has proved robust in deserts (reviewed by Ward
et al. 2013). Coupled with an understanding of phenological differences among
species, the Walter two-layer model has also proved valuable for understanding
different interspecific responses to amount and seasonality of precipitation in
deserts.
11
Plants in Deserts
323
Nurse plant associations are found among many plant families and thus do not
appear to be phylogenetically restricted; however, in deserts the association
strongly benefits establishment of CAM succulent species of the Cactaceae family.
In the establishment period, tender CAM seedlings are sensitive to many environmental and biotic forces that are ameliorated by the presence of the nurse plant. The
nurse canopy reduces direct solar radiation and high temperatures by shading and
attenuates low overnight and winter temperatures. Surface water availability may
also be greater beneath a nurse canopy due to shading or from water supplemented
by hydraulic redistribution. Nurse plants also offer physical protection from herbivory and strong winds, the latter of which may also cause desiccation of young
seedling plants. It is likely that a combination of these factors results in the nurseprotege relationships found among desert cacti and other species.
The saguaro cactus (Carnegiea gigantea) of North Americas Sonoran Desert is
one of the most well-studied protege species of the nurse-protege syndrome
(Fig. 9). Saguaro has multiple nurse species, but the palo verde tree (Cercidium
324
D.R. Sandquist
11
Plants in Deserts
325
References
Eggli U, Nyffeler R. Living under temporarily arid conditions succulence as an adaptive
strategy. Bradleya. 2009;27:1336.
Ehleringer JR, Monson RK. Evolutionary and ecological aspects of photosynthetic pathway
variation. Annu Rev Ecol Syst. 1993;24:41139.
Ehleringer J, Mooney HA, Gulmon SL, Rundel P. Orientation and its consequences for Copiapoa
Cactaceae in the Atacama Desert, Chile. Oecologia. 1980;46:637.
Ezcurra E, United Nations Environment Programme. Division of Early Warning and Assessment.
Global deserts outlook. Nairobi: United Nations Environment Programme; 2006.
Flowers TJ, Colmer TD. Salinity tolerance in halophytes. New Phytol. 2008;179:94563.
Fonteyn PJ, Mahall BE. Competition among desert perennials. Nature. 1978;275:5445.
Fonteyn PJ, Mahall BE. An experimental analysis of structure in a desert plant community. J Ecol.
1981;69:88396.
Kearney TH, Shantz HL. The water economy of dry-land crops, Yearbook of the United States
Department of Agriculture 1911. Washington, DC: Government Printing Office; 1912.
Meigs P. World distribution of arid and semi-arid homoclimates. In: UNESCO, editor. Reviews of
research on arid zone hydrology. Paris: United Nations; 1953.
Mooney HA, Ehleringer J, Bjorkman O. The energy balance of leaves of the evergreen desert
shrub Atriplex hymenelytra. Oecologia. 1977;29:30110.
Mooney HA, Bjorkman O, Collatz, GJ. Photosynthetic acclimation to temperature in the desert
shrub Larrea divericata. Plant Physiology. 1978;61:40610.
326
D.R. Sandquist
Further Reading
Evenari M, Noy-Meir I, Goodall DW. Hot deserts and Arid Shrublands. New York: Elsevier; 1985.
Ward D. The biology of deserts. Oxford/New York: Oxford University Press; 2009.
Whitford WG. Ecology of desert systems. San Diego: Academic; 2002.
Web Resources
http://worldwildlife.org/biomes/deserts-and-xeric-shrublands
http://www.eoearth.org/view/article/168410
12
Matthew J. Germino
Contents
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Alpine and Subalpine Areas Are Valuable for Studying Climate Responses . . . . . . . . . . . . .
Alpine and Subalpine Areas and Vegetation Provide Key Ecosystem
Services in a Warming Climate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Outline for This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
General Description of Alpine Vegetation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
The Environmental Template for Alpine: Soils and Climate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Soils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Microclimate and Energy Balance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Intra-alpine Site Variability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Physiological Responses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Generalized Stress Response and Growth Strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Carbon and Nitrogen Storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
How Do Specific Climate Stresses Occur, and What Are the Physiological Responses? . . . .
Temperature Stress . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
CO2 Availability and Photosynthetic Assimilation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Radiation Stress . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Desiccation Stress . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Linking Microsite, Plant Form, and Physiology in Alpine Plants . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Patterns of Tree Establishment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Ecological Significance of Microclimate Amelioration and Facilitative Interactions . . . .
Future Directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
328
328
329
330
330
334
334
335
343
345
345
345
348
348
349
351
353
353
354
356
360
361
Abstract
Alpine and subalpine plant species are of special interest in ecology and
ecophysiology because they represent life at the climate limit and changes in
their relative abundances can be a bellwether for climate-change impacts.
M.J. Germino (*)
Forest and Rangeland Ecosystem Science Center, US Geological Survey, Boise, ID, USA
e-mail: mgermino@usgs.gov; germmatt@isu.edu
# Springer Science+Business Media New York 2014
R.K. Monson (ed.), Ecology and the Environment, The Plant Sciences 8,
DOI 10.1007/978-1-4614-7501-9_12
327
328
M.J. Germino
Perennial life forms dominate alpine plant communities, and their form and
function reflect various avoidance, tolerance, or resistance strategies to interactions of cold temperature, radiation, wind, and desiccation stresses that
prevail in the short growing seasons common (but not ubiquitous) in alpine
areas.
Plant microclimate is typically uncoupled from the harsh climate of the
alpine, often leading to substantially warmer plant temperatures than air
temperatures recorded by weather stations.
Low atmospheric pressure is the most pervasive, fundamental, and unifying
factor for alpine environments, but the resulting decrease in partial pressure
of CO2 does not significantly limit carbon gain by alpine plants.
Factors such as tree islands and topographic features create strong heterogeneous mosaics of microclimate and snow cover that are reflected in plant
community composition.
Factors affecting tree establishment and growth and formation of treeline are
key to understanding alpine ecology.
Carbohydrate and other carbon storage, rapid development in a short growing
season, and physiological function at low temperature are prevailing attributes of alpine plants.
A major contemporary research theme asks whether chilling at alpine-treeline
affects the ability of trees to assimilate the growth resources and particularly
carbon needed for growth or whether the growth itself is limited by the alpine
environment.
Alpine areas tend to be among the best conserved, globally, yet they are
increasingly showing response to a range of anthropogenic impacts, such as
atmospheric deposition.
Introduction
Alpine and Subalpine Areas Are Valuable for Studying Climate
Responses
The alpine zone contains low-statured, non-arboreal vegetation that is distinct from
lower-elevation, subalpine vegetation, such as forests and occasionally shrub or
grasslands. The highest elevations that vascular seed plants occur at are above 6,000
m in the Himalayas to near 3,000 m lower in high-latitude, maritime-influenced
mountains such as in New Zealand to much lower elevations nearer to polar
latitudes having arctic influences. Much of the area within a particular alpine area
may be unvegetated, particularly at the higher elevations or more exposed sites.
Ground cover may consist of bare rock or soil and with occasional herbs or dwarf
shrubs nestled into features that collect snow. Although alpine areas comprise only
about 3 % of the land on earth, they are distributed across nearly all latitudes and are
highly appreciated for a range of values and ecosystem services they provide to
humans. As a result, alpine areas receive considerable attention given their scarcity.
12
329
The study of plants in the alpine and subalpine has played a foundational role in
fields such as plant ecophysiology and ecology. Elevation gradients and topography
create a pervading physical template for alpine ecosystems. Plant species reach
their low-temperature climate limit in alpine areas and ecosystem effects of climate
are relatively transparent and tractable in the alpine compared to many other habitat
types. The traits of plant species in alpine environments epitomize selection for
stress resistance, versus traits that enhance ruderal or competitive abilities in more
disturbed or temperate growing conditions and complex blends of these strategies
that occur in other habitats. A number of important theories on plant-climate
relationships, such as on microclimate, stress physiology, resource storage, population genetics, and facilitation in plant communities, have roots in classic studies
conducted in alpine environments.
Alpine areas are typically near the upper reaches of mountains, and much of the
global alpine area is in relatively small patches referred to as sky islands
subtended by distinctly different, subalpine ecosystems. Considerable heterogeneity in microclimate, snow cover, vegetation, and soils typically occurs within alpine
areas as an outcome of topographic relief and extreme climate. As a result,
boundary or edge effects and flow between alpine and surrounding ecological
zones are relatively important for alpine ecology (Seastedt et al. 2004). Montane
and particularly alpine areas also provide the opportunity for upward migration and
refugia for species as climate warms (Grabherr and Pauli 1994).
330
M.J. Germino
12
331
summit
SNOW COVER
TOPOGRAPHIC SITE
cu
sh
ridge
ion
snow free
pla
n ts
upper slope
winter snow cover
Geum turf
lower slope
ly
ear
snow
accumulation
d
wbe
sno
late
snowbed
lee exposure
Deschampsia meadow
basin
Carex bog
Salix thicket
stream bed
windward exposure
WIND EXPOSURE
Fig. 1 A classic example of plant community variation attributed by plant or community type or
genus along elevation (topography), wind, and snow cover gradients in the Beartooth Mountains of
Wyoming. Geum is a widespread genus of leafy and rhizomatous herbs commonly called avens,
Deschampsia is a grass, Carex is a sedge, and Salix is willow (Reproduced from Billings and
Mooney (1968), with permission)
332
M.J. Germino
Fig. 2 Representative plant forms of alpine zones. Top left: an herb with cushion form (Phlox)
nested into a rock cranny. Top right: frozen leaves of the leafy herb Erythronium grandiflorum at
sunrise, surrounded by snow in a late-lying snowbed that the shoot had emerged through in the
previous days. Lower: krummholz with flagged stems emerging on Picea engelmannii and Abies
lasiocarpa at treeline in the Medicine Bow Mountains of Wyoming, USA (Photo credits MJ
Germino and W Bowman)
in the Andes. Unique giant rosettes are common in tropical or subtropical alpines,
such as the silverswords of Hawaii (Argyroxiphium), Lobelia of Africa, and
Espeletia of South America (Rundel et al. 1994). Whereas most other alpine
vegetation has small leaves oriented in upright positions, these giant rosettes can
have very large and hairy leaves. In Lobelia, leaves fold over buds to insulate them
at night.
Where trees occur in timberline zones, a frequent pattern with increasing
elevation is that (1) large unforested gaps are found in conterminous forest and
then at higher elevations, (2) trees become islands dispersed within subalpine or
alpine vegetation (meadows), then (3) the timber-like structure of trees is lost and
near true alpine, and (4) any trees present may instead appear in low-prostrate-like
growth forms known sometimes as krummholz (German for twisted wood). In
many mountains, this transition from forest to alpine can occur over many meters or
kilometers, and in others it occurs as a relatively sharp transition, with a change
from forest to alpine occurring within just a few meters. The uppermost elevations
supporting trees in their timber-like form (e.g., several m or taller) are referred to as
timberline, and the uppermost elevations supporting trees in their reduced, often
12
333
334
M.J. Germino
Soils
Alpine soils are generally shallow and not as strongly stratified as lower-elevation
soils that have a longer and more history of weathering, biotic inputs, and stability.
12
335
Where soils in alpine areas are derived of the local geologic substrate, they are
frequently only partially developed and thus have a high proportion of rockiness
and coarse textures. Many alpine areas have soils derived from aeolian inputs (from
air), and local fluvial processes and redistributed (and more weathered) fine soil
particles within basins, generating pockets of deeper and finer-textured soils in
which soil fertility can be partly attributed to upwind or upstream sources. Most
alpine areas have a mosaic of soil conditions that covary with the plant community
variation. In Fig. 1, one might expect a gradient of decreasing soil depth, fraction of
fine particles, and soil fertility from the bottom to top of the triangle. Soil conditions
can be exceptionally patchy in alpine and treeline environments as a result of these
physical processes and feedbacks from plants on the soils beneath them.
Unlike polar tundra, permafrost is not a common and pervasive factor for the
zones of alpine areas that support abundant vegetation. High mountains occasionally have pockets of permafrost at the upper limits of plant life, at depths ranging
from near surface to 0.51 m. Permafrost in the alpine can result from contemporary hydroclimate, or it can be a vestige of previous glaciers or climatic conditions.
Freeze-thaw action causes considerable turbation and forms polygons where coarse
textures are sorted, or frost hummocks, or promotes downhill soil creep called
solifluction that are examples of cryopedogenesis which have significant effects
on plant community structure. Ice crystals that are several cm or longer can
protrude through soil following frost events and cause considerable disturbance to
plant roots.
Where organic matter is present in alpine soils, it is frequently in coarse forms,
which reflects low-temperature inhibition of microbial decomposition processes, in
addition to low inputs of plants. Soil organic matter can range widely among
microsites (up to 550 % of soil mass) and is generally relatively high in the
subalpine and unsurprisingly low on exposed ridges with low plant abundance
and also low C/N. Much of the carbon in alpine ecosystems resides in soil,
reflecting a small standing stock of plant carbon and slow turnover of plant litter
in soil. Alpine soils can have total carbon and nitrogen contents that are similar to
low-elevation ecosystem that also have similar soil depths and textures, but these
nutrients may be more bound in organic forms and are only slowly mineralized in
alpine. Uptake of organic nitrogen, either directly in some cases or more generally
through mycorrhizae (soil fungi attached to roots), is likely key to alpine plants.
Nitrogen content of leaves tends to increase along elevation gradients, reaching
45 % of leaf dry mass for some high alpine herbs as a result of conservation
strategies such as reserve formation, efficient resorption and recovery from
senescing tissue, and accumulation (Monson et al. 2006; discussed further below).
336
M.J. Germino
Fig. 3 Daytime
microclimate temperatures
for cushion plants and leafy
herbs in the alpine, as cited in
Korner (2003)
Air temperature is perhaps the most central climate parameter used to distinguish
the climate of alpine areas, at least at coarse scales, but microclimate relates most
directly to alpine plant ecology.
Microclimate is the temperature, radiation, and wind experienced by plant tissues
such as leaves or flowers. Leaf and plant microclimate is typically very different
from the climate of a site. For example, during the day, temperatures of leaves and
stems can be elevated considerably above the temperatures of the air surrounding
these tissues, particularly when wind speeds are low (Fig. 3). Soil surfaces or leaves
in cushion plants or krummholz can become up to 15 C warmer than the surrounding
air under these conditions. On clear-sky and windless nights, these same leaf and soil
surfaces can become several to 10 C cooler than the surrounding air.
Temperature gradients between alpine cushion plants and the soil and air
temperatures around them are commonly used to demonstrate the concept of
microclimate and to illustrate that plant form can ameliorate the harsh microclimate
to enable optimum growing conditions that could never be appreciated from merely
relying upon on weather station data to predict ecosystem activity.
The degree to which plant microclimate is coupled to site climate and particularly air temperature differs as a function of climate and plant form: cloudiness and
windiness. Both increase the coupling, such that taller plants or plants with small
and sparsely arranged leaves have temperatures that are more similar to the
surrounding air. Plant tissues that are covered with snow tend to have a temperature
of the snow (near 0 C) and they are not subjected to wind or radiation stresses that
prevail above snow. Thus, snow cover is a major factor affecting the climate that
plants actually experience in the alpine.
12
337
Temperature is a unifying factor both for alpine plant ecology and for relating
the interactive effects the main climate factors affecting plants. The ecological
effects of low temperatures in the alpine have been addressed many times, and a
number of studies found correlations of soil-temperature thresholds or minimum
winter temperatures to treeline patterns (Korner 2003, Harsh et al. 2009). Three
energy balance parameters help to relate the alpine climates to plants and their
temperature regime: (1) radiative heat exchange, including both the visible shortwave radiation in sunlight and the long-wave radiation that primarily has thermal
influences; (2) sensible heat exchange, including conductive heat flow from plant
surfaces that are in contact with soil, snow, or water and the more prevalent
convective or wind-affected heat flow from plant surfaces to air; and (3)latent
heat exchange where heat energy is exchanged when water undergoes phase
changes from solid (ice, frost), liquid, and vapor phases.
A key point for plants, particularly alpine plants, is that all three of these modes
of heat exchange can either add or remove heat energy from the plant. When they
have a net effect of adding energy, the plant will be warmer than the surrounding
air, which is nearly always in daytime under sun exposure. It is less well appreciated
but nonetheless important that a net removal of energy by these heat exchange
processes causes plant surfaces to become cooler than the surrounding air. A net
heat loss from plants to the surrounding environment and corresponding cooling of
the plant below surrounding air temperatures typically occurs during night, but can
also occur when high moisture availability results in high transpiration and latent
heat loss.
The degree to which leaf and air temperatures are coupled, and the manner in
which radiative, convective, and latent heat exchanges interact with one another in
regulating plant temperature, can be appreciated from a conceptualization of the
energy balance equation, as follows:
Net radiation Latent heat
Leaf surfacetemperature Air temperature
Convection
The equations for radiative, latent, and convective heat exchange all relate a flux
(of photons, water molecules, or heat) to temperatures with coefficients that translate temperatures into energy units, such as Watts. In a nutshell, this equation tells
us that the difference between the temperatures of leaves or other plant surfaces and
the surrounding air is increased by net radiative (e.g., sunlight) or latent heat
exchange (e.g., transpiration) and is minimized by convective heat change
(wind). The actual energy balance equation, which is solvable, essentially states
that under steady state conditions (i.e., leaves or whatever surface of interest are not
warming or cooling over time and that all components of the energy balance are in
equilibrium), radiative, latent, and sensible heat exchanges must sum to zero.
Metabolically generated heat and heat storage sometimes need to also be entered
into consideration for alpine plants, for the rare cases where electron flow in
mitochondria is not coupled to NADH reduction and for cases where large water
stores occur in succulent plants or large stems.
338
M.J. Germino
Temperature
The low temperatures of alpine areas result from the adiabatic lapse rate, which
refers to the decrease in temperature per increment of elevation as a result of the effect
of decreasing atmospheric pressure (and correspondingly, lower density of molecules
per unit air volume) and pressure-volume-temperature interactions (i.e., PV rRT).
Lapse rates differ regionally, but as simulated they range from 3 C/km to 6 C/km for
relatively humid or dry conditions, respectively (Smith and Johnson 2009). In
addition to these broad elevation gradients in temperatures, there can be more
localized temperature gradients that result from cold-air drainage. Cold air holds
less vapor, and vapor is relatively light among the molecules in air. Thus, cold air is
dense and either tends to settle near ground or drains along the same watershed paths
that water follows. Under clear-sky and windless conditions at night, air temperatures
around low-statured alpine herbs can be 5 C or more below the air temperature at
several m height above ground, as recorded by typical weather stations.
The prevailing climate of a given alpine zone can be challenging to measure, given
the considerable variability in topography associated with alpine areas. Representativeness of available data is a concern. The temperature measured in a radiationshielded weather box or gridded models of temperature at ~1 km pixel resolution that
are parameterized by and designed to predict air temperature at 2 or more m height is
the typical data available. Few weather stations are positioned in a way that can give
temperatures representative of the alpine and subalpine zone of interest here (the
vegetated zone above forest), and many alpine area patches are not well represented
by ~1 km gridded climate models. Thus, considerable uncertainty must exist in our
ability to actually know the temperature or climate that prevails upon alpine plants,
except that plant and soil temperatures are typically near 0 C when snow covered.
Which temperature is of interest? The temperatures most commonly used to
characterize alpine areas are primarily average annual and minimum temperatures,
which are useful in gauging likelihood of snow, but most alpine vegetation is at
least partially covered in snow and so climate during the snow-free season is of
interest. Minimum (nighttime) temperatures during the snow-free growing season
are germane to the majority of alpine plants (except some cushion plants that may
become uncovered in winter or trees protruding above snow). Daytime temperatures during this period will relate most directly to the bulk of physiological
processing. Alpine areas frequently have exceptionally high diurnal temperature
variation. For example, leaf temperatures might increase up to 35 C (e.g., 7 C to
28 C) as air temperatures warm from around 018 C from sunrise to sunset, for
leafy herbs, cushion plants, or krummholz on clear days in early summer or fall. It is
common for alpine plants to be covered with white surface frost before sunrise,
even while air temperatures are reported to be > 0 C by weather stations. With
increasing elevation into the alpine, any surface is likely to warm more above air
temperature during days (Fig. 4).
12
339
Radiation
Solar and long-wave radiation balances in alpine compared to lowland environments may vary according to cloud cover tendencies, and some alpine areas have a
greater incidence of clouds than the surrounding lowlands. The reduction in atmospheric molecules, aerosols, or particulate matter at high elevations results in
greater radiation exchange in alpine areas (Fig. 4), except where alpines occur
within cloud bands (immersion or high elevation).
Shortwave, Solar Radiation
In the absence of confounding factors, sunlight availability increases with elevation
(Fig. 4). Ironically, some of the highest solar radiation levels ever observed happen
to be in continental alpine areas when large cumulonimbus thunderhead clouds
form on otherwise clear-sky days. In this condition, sunlight can reflect off of the
large white cloud walls and add to the already bright direct beam of sunlight. Snow
banks that linger into summer also reflect a considerable amount of sunlight,
resulting in maximum solar radiation (for a surface normal to the sun) in the visible
wavelengths near 3,000 mol m2 s1. The reflected sunlight onto other faces of the
plant adds to this amount. Another significant means by which solar radiation (and
temperature) is appreciably greater in alpine areas occurs when closed, lowerelevation basins (such as in the Great Basin, USA) develop wintertime temperature
340
M.J. Germino
inversions that effectively trap both cold air and haze over lower elevations while
extended high-pressure systems and a corresponding clear-sky and low wind
condition prevail on alpine areas.
The solar radiation intercepted by alpine plants is strongly affected by their leaf
and plant form. Leaves on most species in the alpine tend toward steeply inclined,
upright orientations, which leads to a reduction in the intensity of sunlight
intercepted, at least at midday and at midsummer. The reduction in sunlight energy
intercepted compared to the amount available is a function of the cosine of the leaf
inclination angle (Lamberts cosine law), such that a leaf with a 50 inclination at
45 latitude might intercept less than 1/3 of available sunlight at midday but yet
might increase interception in the morning or evening. Leaf orientation is highly
plastic in many species, and many herbs and short-needled conifers exhibit steeper
leaf orientations in the alpine compared to microsites at lower-elevations or nearerto-forest canopies. On the other hand, some snowbed herbs and pines show less leaf
inclination (e.g., Caltha leptosepala, Pinus flexilis), but these species often have
relatively high physiological tolerance to bright sunlight compared to species like
Abies lasiocarpa that exhibit both sensitivity to sunlight and steep leaf angles
(Germino and Smith 2000). Alpine leaves are also usually relatively thick and
have a lower specific leaf area (cm2/g), which are attributes common in plants of
other sunny and stressful environments (e.g., deserts). Alpine leaves often have leaf
hairs (trichomes) that can impart a light color and thus high albedo.
These morphological adjustments tend to be coordinated with anatomical features that affect sunlight as it propagates into leaves (Smith et al. 1997). Multiple
layers of mesophyll cells are common in leaves of alpine plants and act to increase
photosynthetic capacity per unit leaf area. Differentiation of mesophyll into relatively longer and column-like palisade cells tightly packed toward the epidermis
aid in channeling light deeper into the leaf. Upright leaves in the alpine tend to have
greater isobilateral symmetry, meaning that there is less distinction between the
upper and lower surfaces of leaves than horizontal leaves, such as forbs of temperate mesic environments. In many alpine species that have upright leaf orientations,
palisade cells occur on both sides of the leaf (ab- and adaxial), and this is accompanied by more even distributions of stomata and other features.
Long-Wave Radiation Balance
One of the least well-appreciated but very significant aspects of alpine plant
microclimate is the long-wave radiation (also called infrared or thermal) exchange
between leaves and the environment. The net balance of long-wave radiation
exchange is often negative for leaves and thus constitutes an important cooling
mechanism that can make a habitat with low air temperatures even cooler for plant
surfaces.
All objects emit and absorb long-wave radiation as a function of their temperature, according to the Stefan-Boltzmann equation. Consider a broad leaf of an
alpine herb at night that is oriented horizontally with a temperature of 0 C, and soil
temperatures beneath it are the same temperature. The leaf emits and receives
radiation from both soil and the sky. The leaf and soil emit the same radiation to
12
341
one another, creating a null balance of long-wave radiation exchange for the leaf in
its lower hemisphere. In the upper hemisphere, the leaf is emitting the same amount
of radiation to the sky as toward the soil (300 W/m2 in each direction), but the clear
nighttime sky above the leaf has an effective temperature that can be ~ 50 C that
might equate with only about 150 W/m2 received by the plant from the sky.
Radiation from the sparse molecules in air originates from many km above the
earths surface, from the cool atmosphere. The lower molecular and particulate
density of the air and atmosphere of high-elevation alpine areas further reduces the
incoming radiation relative to lower elevations (Jordan and Smith 1995). The net
outcome is net negative radiation balance at night (150 W/m2) that causes the leaf
to be cooler than air when the energy balance is at steady state. For a leafy alpine
herb with ~2 cm-wide leaves, the leaf might be several degrees cooler than air under
this negative long-wave radiation balance condition if there is little wind. Larger
herb leaves and krummholz shoots can be up to ~10 C cooler under this condition.
Such radiation cooling frequently causes plant, flower, and other surfaces to
exhibit radiation frost at any month of the year, even when site air temperatures
are >0 C, and this is particularly important with the cool nights that occur
throughout the growing season in the alpine. Long-wave radiation balances can
also be negative during day, but incident solar radiation is so large (e.g., 1,000 W/m2)
that physiological impacts of long-wave radiation become significant primarily at
night, particularly clear nights (clouds are much warmer than a clear sky).
Leaf orientation affects net long-wave radiation balances, in addition to interception of sunlight, in ways that greatly affect leaf microclimate and particularly
frost occurrence in the alpine. For example, the horizontal leaf with a 150 W/m2
net radiation balance described above might have a net radiation balance closer to
0 W/m2 in its upright position, if surrounded by other plants or landscape features.
In the case of grass canopies, which contain many leaf blades with relatively upright
orientations, the very top of the leaf blades and canopy is exposed to the sky,
develops a negative long-wave radiation balance, and can cool well below air
temperature, and the resulting frost effects are known to inhibit other species as
they emerge above the grass canopy (e.g., trees in subalpine meadows of Australia
or Rocky Mountains; Ball 1994; Smith et al. 2003). Notably, the upright leaf
orientation in many alpine plants is not always accompanied by other forms of
sunlight avoidance such as the high albedo (low absorbance) of many desert leaves
(e.g., upright but low-albedo leaves of E. grandiflorum in Fig. 2). This may
indicate that long-wave radiation balances are relatively influential for alpine leaf
morphology, which is a prospect that would require further investigation.
Day Length and Seasonality
The strong effects of radiation balance at day and night result in appreciable
differences in the thermal regime of alpine areas at different latitudes. Tropical
alpine areas do not benefit from the additional hours of sunlight and warming that
occur at mid to upper latitudes and instead have more hours of radiation cooling at
night. Cold soil temperatures tend to linger through the entire growing season in
tropic mountain ranges, and there are fewer occurrences of the extreme winter
342
M.J. Germino
conditions and dormancy that prevail at higher latitudes. At higher latitudes, long
days lead to greater intensity of seasonal warming and a sharp transition from
energy-limited growth conditions in early spring (i.e., sunlight or heat-limited) to
water-limited conditions as temperature limitations essentially disappear at midsummer, except for periodic nighttime frosts. Also, whereas lingering snow tends to
insulate many alpine plants from the cool nights of spring, there is much greater
incidence of intense nighttime frost in autumn, as long nights resume at mid to high
latitudes.
12
343
Convection
The aerodynamic shapes of alpine plants, or their tendency to seek microsite
shelter from wind, indicate that wind can be a significant aspect of the climate
(Fig. 2). There is a greater incidence of extreme exposure to winds due to the
landscape prominence of mountains and reduced standing crop and plant canopy
would otherwise impose frictional drag on airflow and protect much of the
vegetation within the canopy from wind speeds. However, some alpine areas can
also be sheltered from wind and wind does not increase with elevation per se
(Korner 2003).
Convective heat exchange is a function of wind speed and the temperature
difference between plant surfaces and air temperature. Alpine plant forms strongly
and often strikingly affect the wind or convection actually experienced by leaves,
particularly for cushion plants or krummholz (Fig. 2). Convective heat exchange is
inversely related to the aerodynamic boundary layer of the leaf, plant, plant canopy,
and, in cases, the tight connection of many alpine plant species to protective
microsite features such as rocks (crannies). The boundary layer can be considered
a buffer zone of air in which air conditions are mutually affected by the leaf or other
surface and the bulk air of the surrounding environment.
An important conceptual consideration for convection is that the energy balance
of plants is affected by boundary layers that are nested into other boundary layers.
For example, narrow cylindrical leaves (e.g., conifers) that have a steep orientation
may have little boundary layer by themselves, but the dense packing of such needles
into shoots, then shoots onto the whole plant, and plants into krummholz mats or
tree islands results in considerable boundary layer resistance each level of
organization has its own contribution to the overall resistance. Leaves of alpine
cushion plants or krummholz may be small and by themselves have small boundary
layers, but they are affected by large boundary layer resistances of the crown and
the soil surface that the foliage is so close to.
Wind can also cause mechanical damage to plants, causing loss of shoots or
leaves or reproductive parts. During winter, snow that remains on the ground for
extended periods can have crystal metamorphosis that generates granules with
sharp edges, which abrade exposed plant tissue when blown across the landscape
at high speeds (Smith et al. 2003). The dense clustering of leaves or stems into the
individual plant or canopy (e.g., tree islands) can help optimize other aspects of
wind, by reducing wind enough to trap snow in ways that protect the buried plant
from extreme temperatures or snow abrasion.
344
M.J. Germino
landscape and vegetation. As shown below, the degree of alteration is not trivial and
the resulting ecological feedbacks have important outcomes for alpine ecology.
Trees affect the microclimate around themselves in several ways, by providing
shade from sunlight, which affects microclimate and photosynthesis of the shorter
neighboring vegetation (Ball 1994; Germino and Smith 2000). As an example of
microclimate effects, in mid to upper latitudes, it is common to see triangular areas
to the north (northern hemisphere) or south (southern hemisphere) that retain frost
or snow when the rest of the landscape has melted, often for days or weeks.
At night, tall-statured trees increase the amount of long-wave radiation incident
upon the plants and soil surrounding them, which increases surface temperatures,
particularly minimum daily temperatures at night. Trees also affect wind flow,
providing a bluff body effect that can cause appreciable snow drifting in their lee
that inspired the term snow glades. The snow drifts from a single tree island can
extend nearly 100 m in length across subalpine or alpine-like meadows and can reach
depths of 10 m or so. Furthermore, snowdrifts can endure months following the melt
out of the surrounding landscape. For example, in Western North America, most
snowmelt in the portion of alpine areas having relatively abundant plant cover occurs
by late May through June, occasionally into July, whereas large snowdrifts even
within the lower portions of the timberline zone may persist into or through August,
and in some years may even have some of the snow bank present when snowpack
begins to accumulate between September to November. Snow cover insulates plants
from prevailing climate, provides an important source of water during the growing
season, and stimulates microbial activity such as the pathogenic Herpotrichia that
smothers vegetation in late-lying snow banks with a coating that resembles tar.
Topographic features such as hills or cliffs can have some of the same solar
shading, long-wave enrichment, and bluff body and snow drifting effects as tree
islands but at a greater range of scales. Additionally, the drainage effects of
topography on both water and cold air can have very large effects on the microclimate patterns of entire alpine landscapes. For demonstration, a useful exercise is to
consider what the coldest location within an alpine landscape might be, given the
microclimate and energy balance considerations described above. It would likely be
a microsite that has high sky exposure (i.e., is distant from tree islands or large cliffs
that occlude the view of sky) that is also in a closed (i.e., no outlet draining) basin
with slopes that are steep enough to drain cold air to the microsite but not so steep as
to block the sky view. In the absences of cloud cover that would moderate the longwave radiation balance or strong winds and mixing of air, air and surface temperatures near ground could easily be 1020 C cooler than microsites with opposite
conditions (e.g., on a ridge with sky-occluding features). This cold spot would also
receive high sunlight during days and probably would have relatively warm midday
conditions, and snow drifting and soil moisture would also differ. These drainage
and sky exposure considerations can help explain phenomena such as inverted
treelines, in which slopes above alpine-like (or subalpine) meadows are forested.
As demonstrated below, the physiology of alpine plants is very much linked to the
combinations of these different microclimate factors, perhaps more so than to any
single factor itself.
12
345
Physiological Responses
Generalized Stress Response and Growth Strategies
The environmental challenge for plants in an alpine environment is to rapidly
utilize available snowmelt for growth during a short and/or cool growing season.
Research has asked how plant uptake of carbon and soil resources and growth
processes are impacted or adapted to short and cold alpine growth conditions. The
short growth season and prospects for alternation of favorable and less favorable
growing seasons (e.g., very short growing season or one with extended water
deficit) have led to three key aspects of plant adaption to the alpine.
First, anticipatory development is common in many herbs, in which buds are
preformed in the fall prior to winter dormancy, enabling rapid development upon
spring or summer snowmelt. Unlike plants from temperate environments, vegetative and reproductive growths tend to be synchronous, although preformation can
occur in either type of meristem. The prevalence of bud preformation should have
the effect of decoupling growth of alpine herbs from the weather prevailing in any
given year. Reliance on bud preformation limits the ability of alpine plants to adjust
their development to current conditions, however.
Second, rapid shoot emergence is subsidized by carbohydrates and nitrogen
acquired in previous growing seasons and stored in large root systems or belowground storage organs. With these advantages, species such as marsh marigold
(Caltha leptosepala) and several buttercups (Ranunculus sp.) are well known to
begin development in the relatively low-light and near-freezing conditions under
snow, frequently developing through snow and completing much of their life
history nearly in contact with the snow retreating from around them (Billings and
Mooney 1968).
Third, alpine herbs can exhibit relatively high rates of resource uptake when
conditions are optimal, and they are furthermore uniquely able to sustain uptake and
growth under cool conditions that characterize the growing season. Interestingly,
the highest elevation herbs tend to have a leafy and not cushion physiognomy and
slow growth associated with it, indicating selection upon them for rapid capitalization of growth opportunity at the highest reaches of plant life.
346
M.J. Germino
range of other habitats. The storage pools are quickly depleted upon release from
winter dormancy, when allocation to rapidly expanding new tissue and to respiration exceeds photosynthetic gain. The importance of this translocation-storage
mechanism is evident for snowbed herbs like Caltha leptosepala, which do not
exhibit appreciable storage formation in microsites where growing season length is
truncated by deep snow banks but do accumulate sugar and starch where growing
seasons are longer (reviewed in Billings and Mooney 1968). Seasonal redistribution
of carbon from shoots to roots has also been observed in evergreen plants, although
their foliar carbon content can also increase as plants acclimate to the onset of
drought and particularly winter cold. Carbohydrates also have direct roles in stress
responses; simple and complex sugars (e.g., fructans and raffinose) correspond well
with acclimation to chilling in leaves of Dactylis glomerata and other alpine herbs
(Monson et al. 2006). Sugar and other osmotic compounds decrease the temperature
required for ice formation (antifreeze) and decrease the tendency for water to be lost
from cells, thereby protecting against desiccation.
There has been considerable emphasis on evaluating nutrient pool sizes, especially of carbohydrates in their starch or sugar form, as a means for identifying
processes limiting productivity of alpine or treeline species. Many studies have
revealed increases in carbohydrates or nitrogen content per unit leaf area with
increasing elevation into alpine or treeline zones, in herbs or trees (Korner 2003).
Many of these studies found greater concentrations of carbohydrates at higher
elevations, leading to the suggestion that chilling at treeline does not limit carbon
uptake in trees, but rather their ability to use carbon for growth processes (Korner
1998). These studies relied on estimates of the percent of dry mass that was
available carbohydrate, i.e., nonstructural or mobile sugars or starch in leaves
or stems and occasionally roots. Although it is convenient (and common) to view
carbohydrate concentrations as if they are merely a passive outcome of carbon
sources (photosynthesis) and sinks (growth, respiration, and losses through root
exudation or tissue shedding; Ryan 2011), active regulation of carbohydrate pools
is reflected in variation in concentrations of carbohydrates among alpine and
subalpine plants (Bansal and Germino 2008; Wiley and Helliker 2012). The
concurrence of growth reductions and elevated stores of carbon or nitrogen may
result from an inability to use the resources for growth, but growth may also be
reduced to ensure the formation of the reserves. Strategies like this might be
expected for long-lived perennials in which rare years of very poor net carbon
flux might select for reserve formation abilities, and active reserve formation could
certainly be stimulated by the same factors that directly affect growth and all carbon
source and sink processes.
Active storage creates reserves at the expense of growth or other processes,
whereas passive storage is accumulation with no apparent cost to the plant. Sugars
of many alpine plants and specialized molecules such as cyclic polyols (cyclitols)
accumulate following shoot expansion, while photosynthesis is at seasonal maximum, following a pattern indicative of storage formation (Monson et al. 2006). In
alpine herbs of the Caryophyllaceae, cyclitols confer protection against late-season
drought stress (and probably also nighttime freezing) and their concentrations
12
347
P.menziesii at 2450m
P.menziesii at 3000m
A.lasiocarpa at 2450m
A.lasiocarpa at 3000m
2
r 2 = 0.42
2
3
Photosynthesis:Respiration
Fig. 5 Relationship of nonstructural carbohydrates, specifically starch, and whole-shoot photosynthesis and respiration in the short-term history of tree seedlings planted across the full breadth of
a treeline ecotone (Regraphed data from Bansal and Germino (2008)). Sugars were not related to
photosynthesis: respiration, and starch is generally considered a storage form of carbohydrate and
allocation to growth was not an appreciable aspect of the carbon balance during the time increments
evaluated. With more than half of the variability in % starch unexplained by the passive balance of
photosynthesis-respiration, it appears likely that active regulation such as reserve formation may be
occurring. Species were Pseudotsuga menziesii, which does not normally occur at 3,000 m, and
Abies lasiocarpa, which normally spans the full alpine gradient. The values shown are the means
from different sampling dates spanning the entire snow-free growing season
348
M.J. Germino
12
349
chilling to growth. Many alpine plants, specifically herbs, express a high capacity to
sustain resource uptake and growth at very low temperatures. Snowbed herbs,
known as geophytes, are frequently observed sprouting and beginning stem elongation under snowpack, where temperatures are 0 C (or less, due to nighttime
cooling of the surface). Snowbed herbs can achieve maximal photosynthesis within
minutes of having been thoroughly frozen (Germino and Smith 2000). In the
subalpine, seeds of conifers such as Picea engelmannii and Abies lasiocarpa
commonly germinate and have cm of growth while imbedded in snow banks that
persist into summer growing season, clearly utilizing their carbon reserves and
growing at temperatures near 0 C though seedling establishment may not result.
Freezing, specifically the formation of ice, occurs at lower temperatures in some
alpine plants as a result of freezing point depression or supercooling. By withdrawing water or adding osmotically active molecules derived of carbohydrates or other
ions, the temperature required to cause the apoplast or symplast to freeze can be
decreased considerably, by a few to 20 C or more. Plants that supercool withdraw
nucleating agents, thereby inhibiting ice formation to very low temperatures.
Membrane flexibility can also be adjusted by the fraction of unsaturated C-C
bonds in the lipids comprising the plasma membrane, allowing the plant to avoid
disruption and leakage across the plasma membrane or cell wall. Raffinose and
other carbohydrates can adhere to and stabilize membranes leaves undergoing
freeze-thaw cycles. Rapid dissolution of emboli in xylem elements that become
cavitated during freeze-thaw cycles is also a key adaptation.
The high diurnal fluctuation in temperatures of the alpine are associated with low
nighttime minimum temperatures. Nighttime frosts affect flowering of alpine and
subalpine herbs (Inouye 2008). Experimental enrichment of long-wave radiation
and corresponding increases in nighttime temperature have generated significant
changes in alpine plant communities and in flowering and species shifts (Harte
and Shaw 1995).
350
M.J. Germino
12
351
Radiation Stress
Radiation has a number of positive and negative effects that can appear paradoxical.
The positive long-wave radiation balance warms leaves in the cool alpine, but the
more common negative long-wave balance cools minimum temperatures in an
already cool environment. Visible shortwave (solar) radiation drives photosynthesis
and warms leaves, but can also cause photochemical problems (Ball 1994). Ultraviolet radiation causes photochemical damage by causing somatic mutations of
DNA, but there is no evidence that alpine plants are damaged more by UV than
lowland plants. The minimal impact of UV may be due to effective screening of UV
352
M.J. Germino
at the epidermis, by solutes and by pigments such as the red anthocyanin that is
common in alpine and most sunny habitats.
The basic response of photosynthesis to sunlight in alpine plants generally
follows attributes of leaves adapted to sunny environments, which includes the
anatomical and biochemical traits described above. The photosynthetic response to
step changes in sunlight usually reveals that photosynthesis saturates at relatively
high sunlight levels in alpine plants.
High sunlight levels can also have negative effects on photosynthesis, leading to
a condition known as photoinhibition (Ball 1994). Photoinhibition refers to the
light-dependent reductions in photosynthesis that tend to occur when plants are
subject to other stresses, particularly low temperatures. When low temperatures
cause a reduction in enzymatic activity, the processes in which recently assimilated
carbohydrates are reduced (i.e., the dark reactions, Calvin cycle) and their export
from chloroplasts for use in growth become slower. The enzymatic dark reactions
are the primary consumers of the reducing equivalents (ATP, NADH) produced
in the light reactions, the so-called Z-scheme of chlorophyll reaction centers and
transmembrane proteins (thylakoid membranes in chloroplasts) that produce ATP
and NADH from the energy derived from sunlight. The outcome is a net imbalance
of supply and usage of sunlight excitation energy supply in photosynthetic carbon
reduction (supply of ATP and NADH). Under these conditions, the electron transport chain of proteins becomes so reduced (saturated with electrons) that it can no
longer accept electrons from photosystem II, where the sunlight energy harnessed
from the chlorophyll antennae is used to split water and elevate the energy level of
the resulting electrons such that they are able to flow through the electron transport
toward the end products ATP and NADH. The excess excitation energy in the
photosystem II complex can be dissipated through reradiation as chlorophyll
fluorescence, which is directly measurable but is not considered a significant
adaptive mechanism. The excess excitation energy can also be safely
dissipated back into the chlorophyll antennae complex, provided that compounds
on the thylakoid membranes known as xanthophylls are in a de-epoxidated
molecular configuration. This safe dissipation is a reversible condition known as
non-photochemical quenching (photochemical quenching leads to ATP and NADH
production). When sunlight energy harnessed by leaves cannot be adequately
dissipated by chilling-inhibited dark reductions or by non-photochemical
quenching, the excess excitation can lead to photooxidative damage and ultimately
nonreversible (or slowly reversible) damage and reduced photosynthesis and
growth.
The photochemical challenge for alpine plant life has been the subject of a
number of studies that have served to illustrate the ecological relevance of molecular processes in the chloroplast or mesophyll. The reversible downregulation of
photosynthesis is evident for many alpine species in diurnal photosynthesis or
chlorophyll fluorescence patterns. Non-photochemical quenching via fluorescence
or concentrations of xanthophylls in sunny microsites do not show a clear increase
with elevation, suggesting that although xanthophyll configuration may more
consistently allow non-photochemical dissipation of sunlight energy, alpine plants
12
353
may not necessarily have greater capacity for this protective means (Germino and
Smith 2000). One of the most important means for avoiding photoinhibition in
alpine species may simply be the ability for photosynthesis to continue at low
temperatures, which sustains the consumption of photochemical energy and minimizes excess absorbed excitation energy. There is greater evidence for alpine plants
to have more antioxidants as a means for mitigating damage associated with
nonreversible impacts of photoinhibition (Germino and Smith 2000). Reports of
state transitions (spatial changes between reaction centers that redistribute and
balance excitation energy), photorespiratory capacity, the Mehler reaction (electrons from water ultimately reduce peroxide), and other processes have been
evaluated in alpine plants, creating an intriguing body of research that continues
to illustrate the ecological role of sunlight in the alpine. Low-temperature
photoinhibition illustrates the manner in which multiple stresses can interact in
alpine environments, in ways that are not simply additive.
Desiccation Stress
Evapotranspiration can be modeled using a diffusion or Ficks law analog to
that presented above for photosynthesis, as the product of the leaf-to-air
vapor concentration (or partial pressure) gradient and of the conductance of the
leaf-to-water vapor transport into the bulk air. The high rate of molecular diffusion
in the low atmospheric pressure of the alpine causes high leaf conductance to water
vapor (just as it enhances stomatal conductance to CO2). When combined with high
leaf-to-air temperature and thus vapor gradients for cushion plants, krummholz, or
large-leaf herb, transpiration potential is high (Smith and Johnson 2009).
The osmotic pressure of alpine leaves tends to be less (having more solutes)
than lower-elevation environments, which may suggest adaptation to drier
conditions but could also be linked to active regulation for carbohydrate reserve
formation or freezing avoidance. A fundamental difference between semiarid
basins and alpine areas, however, is the relatively greater water supply typical of
alpine areas.
354
M.J. Germino
12
355
Emergents
Seedlings
80
% SURVIVAL
80
40
20
60
40
20
(8,11,3)
N
(9,8,2)
80
60
40
20
Emergents
Seedlings
Saplings
% SURVIVAL
60
% SURVIVAL
Emergents
(5,11,1)
(8,1,2)
(5,8,2)
(9,10,3)
(7,8,2)
(6,8,2)
Prevailing wind direction
Fig. 6 Tree seedling survivorship around tree islands in a treeline ecotone of Wyoming, USA, of
Picea engelmannii and Abies lasiocarpa. Each axis is a cardinal direction, with the origin being the
canopy edge of a tree island. Increments along each axis are annual survival. The outer line shows
100 % survival. Blue shows survivorship of newly germinated (emergent) seedlings in their first
summer after germinating; gray shows seedlings are plants in the 2nd to 5th year of growth, and
saplings are plants older than 5 years but less than 20 cm tall. Numbers in parentheses are sample
size for emergents, seedlings, and saplings, respectively (Regraphed from data reviewed in
Germino and Smith (2000))
sunlight generates symptoms of low-temperature photoinhibition, and the correlation of this proposed mechanism to establishment implies a role for carbon limitation to tree establishment.
Greater carbohydrate concentrations in trees near treeline has prompted the
proposal that the treeline environment poses more challenges to carbon use rather
than uptake (Korner 1998), but young tree seedlings are faced with needing to gain
356
M.J. Germino
manifold increases in carbon from year to year in order to have normal growth and
development and could therefore conceivably have greater carbon limitation
(Bansal and Germino 2008). Depletion of carbohydrates to near 0 % of dry mass
has not been observed in tree seedlings at treeline, but positive associations among
survivorship, photosynthesis, and carbohydrate concentrations have been
established, suggesting tenability of the hypothesis but a need for evidence other
than carbohydrate concentrations. Alternative hypotheses for tree seedling affinities
for tree islands, such as lack of availability of the appropriate ectomycorrhizae for
tree seedlings in herbaceous meadows, or altered soil properties near trees, were
explored by reciprocal soil transplanting but yielded no compelling support.
12
357
Seedling/Sapling
Flagged Tree
Forest Tree
15 m
Wind Direction
Treeline
Timberline
Krumholz Mats
Intact Forest
Snow
Glade
Ribbon Forest
Flagged trees
with mats
Mats only
Fig. 7 Gradient in tree abundance, form, and corresponding seedling establishment from forest to
treeline (left to right), in this case for an east-facing slope with conifers in the Rocky Mountains,
USA. By extension, this schema could also show forest development over time for a fixed location
(right to left; from Smith et al. 2003, with permissions)
Fig. 7 illustrates that greater seedling establishment occurs in the vicinity of flagged
trees that are large and dense enough to ameliorate the nighttime frost and bright
sunlight on the ground around themselves. The snow glade on the left results
from exceedingly deep snow drifts. From right to left, the individual conifer, as
krummholz, ameliorates its own microclimate, but wind shear and damaging winter
snow abrasion strip leaf cuticles, and leaf desiccation and tissue loss inhibit any
upward growth. The krummholz (Fig. 2) form results from a wind-pruning mechanism whereby sharp snow crystals blown across the surface of the snow (saltation)
abrade the cuticle that would otherwise inhibit desiccation. In most springs, red
foliage on the top of krummholz indicates where previous growth was not protected
by the snow drift that forms in the dense foliage (and thus boundary layer) of
the krummholz. As an aside, the factors sculpting the form and thus height growth
of trees at treeline indicate that summertime growth processes (e.g., carbohydrate
storage) need to be considered in light of other processes regulating plant
stature. Regardless of the mechanism affecting height, tree height itself is a major
factor affecting the surrounding landscape. In the next row left (Fig. 7), when
meristems are able to grow rapidly enough in a good year to get meristems
above the saltating snow zone, then flagged stems of greater stature occur.
Mutual microclimate amelioration among krummholz/flag plants can then occur
358
M.J. Germino
12
359
(measured at 1 m height) during day and 8 C cooler than air at night, leading to leaf
temperatures below 0 C on 70 % of nights during its growing season. In sharp
contrast, leaf temperatures of E. grandiflorum deviated only 24 C from air
temperature and had temperatures <0 C on less than 40 % of nights during the
growing season. Interestingly, minimum leaf temperatures occurred not before
sunrise as is commonly thought for all plant life, but instead occurred after sunrise
for C. leptosepala. This appeared to be the result of sustained frost formation before
sunrise and even in the morning twilight while the sun was below the horizon,
followed by very large latent heat losses when thick layers of frost melted and
evaporated within minutes of very bright sunlight exposure.
Differences in leaf temperature between the species are attributable nearly
equally to differences in microclimate air temperature and their differences in
leaf form, as determined by measuring leaves in different orientations and by
experimentally altering leaf angles. Compared to the broad and flat leaves of
C. leptosepala, the slender, upright leaves of E. grandiflorum have higher nighttime
temperatures because their net long-wave radiation balance is greater (exchanging
radiation with surrounding plants and topography instead of sky), their convective
heat exchange is greater due to narrower leaves (upright leaves also more efficiently
drain away the air that cools next to them), and leaves are elevated into warmer air
layers as a result of the steep inclination from horizontal. Eliminating these
morphological advantages leads to up to 6 C cooler leaves.
These snowbed species exhibited a high resistance to frost. No difference in
carbon gain occurred following nights with or without frosts in E. grandiflorum, but
a 35 % decrease in photosynthesis for C. leptosepala for several hours in the
mornings following its more severe blend of chilling and bright light. Both species
exhibited about 8 % reductions in their sunlight-use efficiency from morning to
midday, reflected in decreases in quantum yield (mol/mol of CO2 gained per photon
of sunlight absorbed, at low sunlight) and by chlorophyll fluorescence. Such losses
in photosynthetic efficiency are not expected to be significant when leaves are
saturated with sunlight, and instead the lower post-frost carbon gain of
C. leptosepala corresponds with the greater mutual shading in its crown and canopy
and reports that it can proceed through a growing season without appreciable
carbon reserve formation. Nonetheless, chlorophyll fluorescence did not reveal
that either species had any unusual ability to utilize non-photochemical pathways
of dissipating bright sunlight even while leaves were nearly 0 C in the morning,
and photochemical damage was never evident. Instead, a high capacity to sustain
productive use of sunlight energy for photosynthesis, thereby avoiding excess
sunlight energy absorption, allows these herbs to occupy one of the severest
combinations of environment and life history for chilling and bright sunlight in
nature.
The restriction of C. leptosepala to the wet but cold depressions may be linked to
its hydraulic requirements and may be further enabled by its high leaf elasticity for
enduring freeze-thaw events, and ability to withstand inundation and the associated
hypoxia, as could be revealed with future research. These types of research efforts
may help reveal trade-offs in temperature resistance and water requirements that
360
M.J. Germino
can help explain species niches and landscape patterning of communities (Fig. 1)
as they have resulted from recent climate and may change with the onset of novel
climates and combinations of temperature and water.
Future Directions
With the specter of global warming, alpine ecosystems are increasingly valued as
bellwethers for early and relatively unconfounded impacts of global warming
(Smith et al. 2009; Pauli et al. 2003). Low-elevation ecosystems tend to be more
impacted by the complex interactions of disturbances, invasive species, and species
change resulting from biotic factors such as competition and pathogens, for examples. Furthermore, the broader distribution of the alpine biome across latitudes and
continents should provide a basis for comparison and generalization for warming
impacts. Alpine environments can better serve such a role if accurate predictions
can be made for where and when vegetation change is likely, within and among
alpine areas. These predictions would be possible with an understanding of climatic, landscape, and biotic factors that increase or decrease the vulnerability to
climate change. Furthermore, predicting where and when the change is likely to
occur will help in devise appropriate monitoring systems that are key for adaptive
management. For example, the alpine-treeline ecotone is considered a bellwether
for climate impacts, but how and where should change be evaluated? Young tree
seedling abundances at alpine-treeline may be a relatively sensitive indicator of
incipient change, considering that the tight coupling of tree seedling establishment
in alpine-treeline ecotones to long-wave radiation balance, nighttime frost (and thus
greenhouse effect), and annual climate variability contrasts with older treeline
trees, in which fewer changes are apparent even over centuries of climate variability. Considerations like this help identify demographic stages, particular species,
and points within the landscape that might be more likely to express change that
portends larger ecosystem and landscape transformations. Notably, tree seedling
establishment has increased in many alpine-treeline ecotones globally (Harsh
et al. 2009), but there is considerable and unexplained variation about this tendency
among mountains. Vegetation management tends to occur at scales closer to a
particular alpine site, and an understanding of why the variability occurs will enable
translation of the information in broad-scale vulnerability assessments back to the
land management decisions that are the fulcrum for human adaptation to climate
change.
How can the information needed to enable this cross-scale understanding be
provided? Systematic and uniform sampling across mountain ranges, globally, such
as in the GLORIA project (Pauli et al. 2003), may provide key data for
intercomparison, meta-analysis, and synthesis. Second, identifying where generalizations about the mechanisms governing alpine and treeline responses to climate
(i.e., uphill advance of forest into alpine) as they vary globally can or cannot be
made is a key step toward scaling up information from the many plot- or single-site
studies (Smith et al. 2009). This is a direction pioneered by Ch Korner that is
12
361
challenging, but the data and information needs are feasible and will lead to key
insights.
At the landscape scale, a promising research frontier is on how connectivity for
movement of species or genes influences alpine and treeline change, i.e., biogeographic and evolutionary controls. At the ecosystem level, alpine areas are increasingly influenced by atmospheric deposition of nitrogen and particularly dust, and
how will these changes modify alpine responses to warming? At the plant community level, can concepts like positive, facilitative associations among species be
used to predict local shifts in assemblage? Will the emergence of novel climate
conditions affect the vulnerability of alpine areas to upward migration of exotic
species as a result of their dispersal advantages and opportunism? At the organismal
level, the precise manner in which alpine climates limit plant species, such as trees,
is not resolved. There is considerable debate on whether carbon stores, i.e., mobile
carbohydrates, can be used to identify rate-limiting processes for treeline change,
which illustrates just one frontier in the broader quest to understand physiological
limitation as it applies to understanding plants at their climate limit. Ecophysiological theory suggests that plants operate such that their growth is not limited so
strongly by any one particular factor or process, and the compensatory responses
that generate balance can occur throughout the whole plant. Research addressing
limitation has tended to emphasize a select few processes that we are poised to
measure given available instrumentation, and the assessments are rarely at the unit
of selection on the landscape: i.e., the whole plant. Application of molecular tools,
such as high-throughput genetic approaches to characterizing transcriptomes and
thus the full breadth of how alpine plants respond to changes in their environment,
is a likely path forward.
References
Ball MC. The role of photoinhibition during tree seedling establishment at low temperatures. In:
Baker NR, Bowyer JR, editors. Photoinhibition of photosynthesis from molecular mechanisms
to the field. Oxford: BIOS Scientific; 1994. p. 36576.
Bansal S, Germino MJ. Carbon balance of conifer seedlings at timberline: relative changes in
uptake, storage, and utilization. Oecologia. 2008;158:21727.
Billings WD, Mooney H. The ecology of arctic and alpine plants. Biol Rev. 1968;43:481529.
Billings WD. Adaptations and origins of alpine plants. Arct Alp Res. 1974;6:12942.
Bowman WD, Seastedt T, editors. Structure and function of an alpine ecosystem: Niwot Ridge,
Colorado. New York: Oxford University Press; 2001.
Callaway RM, Brooker RW, Choler P, Kikvidze Z, Lortiek CJ, Michalet R, Paolini L, Pugnaire FI,
Newingham B, Aschehoug ET, Armasq C, Kikodze D, Cook BJ. Positive interactions among
alpine plants increase with stress. Nature. 2002;417:8448.
Cordell S, Goldstein G, Mueller-Dombois D, Webb D, Vitousel PM. Physiological and morphological variation in Metrosideros polymorpha, a dominant Hawaiian tree species, along an
altitudinal gradient: the role of phenotypic plasticity. Oecologia. 1998;113:18896.
Ellison L. Subalpine vegetation of the Wasatch plateau, Utah. Ecol Monogr. 1954;24:89184.
Germino MJ, Smith WK. Differences in microsite, plant form, and low-temperature
photoinhibition in alpine plants. Arct Antarct Alp Res. 2000;32:38896.
Grabherr G, Pauli MGH. Climate effects on mountain plants. Nature. 1994;369:448.
362
M.J. Germino
Harsh MA, Hulme PE, McGlone MS, Duncan RP. Are treelines advancing? A global metaanalysis of treeline response to climate warming. Ecol Lett. 2009;10:10409.
Harte J, Shaw R. Shifting dominance within a montane vegetation community: results of a climatewarming experiment. Science. 1995;267:87182.
Inouye DW. Effects of climate change on phenology, frost damage, and floral abundance of
montane wildflowers. Ecology. 2008;89:35362.
Jordan DN, Smith WK. Energy balance analysis of night-time leaf temperatures and frost
formation in a subalpine environment. Agr Forest Meteorol. 1995;77:35972.
Korner C. A re-assessment of high elevation treeline positions and their explanations. Oecologia.
1998;115:44559.
Korner C. Alpine plant life: functional ecology of high mountain ecosystems. 2nd ed. Berlin:
Springer; 2003.
Leuschner C. Are high elevations in tropical mountain arid environments for plants? Ecology.
2000;81:142536.
Lutz C, editor. Plants in alpine regions: cell physiology of adaptation and survival strategies.
Wien/New York: Springer; 2012.
McDowell NG. Mechanisms linking drought, hydraulics, carbon metabolism, and vegetation
mortality. Plant Physiol. 2011;155:10519.
Monson RK, Rosenstiel TN, Forbis TA, Lipson DA, Jaeger III CH. Nitrogen and carbon storage in
alpine plants. Integr Comp Biol. 2006;46:3548.
Nagy L, Grabherr G. The biology of alpine habitats. New York: Oxford University Press; 2009.
Pauli H, Gottfried M, Reiter K, Grabherr G. High mountain summits as sensitive indicators of
climate change effects on vegetation patterns: the multi summit-approach of GLORIA
(global observation research initiative in alpine environments). Adv Glob Chang Res.
2003;9:4551.
Rundel PW, Smith AP, Meinzer FC, editors. Tropical alpine environments: plant form and
function. Cambridge: Cambridge University Press; 1994.
Ryan MG. Tree responses to drought. Tree Physiol. 2011;31:2379.
Seastedt TR, Bowman WD, Caine TN, McKnight D, Townsend A, Williams WM. The landscape
continuum: a model for high-elevation ecosystems. BioScience. 2004;54:11121.
Smith WK, Johnson DM. Biophysical effects of altitude on plant gas exchange. In: De la
Barrera E, Smith WK, editors. Biophysical plant ecology: perspectives and trends. Mexico:
Universidad Nacional Autonoma de Mexico Press; 2009. p. 25780.
Smith WK, Vogelmann TC, Bell DT, DeLucia EH, Shepherd KA. Leaf form and photosynthesis.
BioScience. 1997;47:78593.
Smith WK, Germino MJ, Hancock TE, Johnson DM. Another perspective on altitudinal limits of
alpine timberlines. Tree Physiol. 2003;23:110112.
Smith WK, Germino MJ, Johnson DM, Reinhardt K. The altitude of alpine treeline: a bellwether of
climate change effects. Bot Rev. 2009;75:16390.
Terashima I, Masuzawa T, Ohba H, Yokoi Y. Is photosynthesis suppressed at higher elevations
due to low CO2 pressure? Ecology. 1995;76:26628.
Tomback DF, Arno SF, Keane RE, editors. Whitebark pine communities: ecology and restoration.
New York: Island Press; 2001.
Wiley E, Helliker B. A re-evaluation of carbon storage in trees lends greater support for carbon
limitation to growth. New Phytol. 2012;195:2859.
13
Kim M. Peterson
Contents
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Boundaries of the Arctic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Desert and Tundra . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Arctic Flora . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Plant Life in the Cold . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Permafrost . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Roots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Nutrients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Moisture and Vegetation Patterns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Geomorphic Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Cryoturbation, Needle Ice, and Other Soil Disturbance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Arctic Climate Change . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Future Directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
364
366
366
369
370
377
378
380
381
382
385
386
386
388
Abstract
363
364
K.M. Peterson
Introduction
The Arctic is a cold treeless expanse of plains, hills, and mountains, including the
northernmost parts of continental Eurasia and North America and numerous highlatitude islands, the largest being Greenland. Collectively these lands surround the
Arctic Ocean. Even though the Arctic Ocean is variously ice covered, like other
oceans it ameliorates climate, reducing extremes of temperature. Despite these
maritime effects, the Arctic is cold and climatically dry. Low temperatures and
13
365
limited heat resulting from low solar angles in summer and darkness in winter keep
the Arctic frozen much of the year. The sun seems to be forever rising or setting
during the brief growing season and vanishes all together for extended periods
during Arctic winters. Water is frozen much of the year creating potential physiological drought, and precipitation is generally low throughout the year. Despite a
lack of Arctic precipitation associated with persistent polar high pressure, locally
moist or even wet habitats are common throughout the Arctic.
Low temperatures and a general lack of heat profoundly affect the ecology of
Arctic plants. Arctic plants face a host of challenges, including freezing temperatures, short growing seasons, limited soil fertility, episodic herbivory, and low
pollinator frequencies. As a result the Arctic flora is small relative to other ecosystems and represents the end of a latitudinal gradient in floristic diversity that begins
high in the tropics and declines to a minimum in the Arctic.
Arctic plants share requirements for light, carbon dioxide, mineral nutrition, and
water common to all plants, and they must be able to meet these requirements
within the unique constraints imposed by the Arctic. Few of the species able to
persist in the Arctic are restricted entirely to Arctic ecosystems. The geographic
distributions of many Arctic plant species extend outside the Arctic to high mountains, bogs, or boreal landscapes, and plant species confined to the Arctic often have
close relatives in alpine or boreal areas. This is to say that many plant species found
in the Arctic can, and often do, grow well outside of the Arctic, but the reverse is not
true, i.e., relatively few plants found outside the Arctic are able to grow and persist
in the Arctic. The Arctic environment is a selective filter, admitting a small flora,
requiring plants to tolerate short cold growing seasons and long frozen winters.
Despite decades of research, many questions of climate change and potential
effects upon the Arctic remain unanswered. Arctic ecosystems encompass a broad
diversity of habitats: Deserts, semideserts, ice caps, glaciers, rock fields, dry
tundras, moist tundras, wet tundras, shrublands, heaths, bogs, marshes, salt
marshes, and aquatic communities are part of the diversity found in the Arctic.
There is no single Arctic vegetation, but a matrix of distinct environments with
distinct vegetational assemblages, each with differing susceptibilities to environmental change (Crawford 2008).
In Arctic landscapes, the magnitude of microhabitat distinctions is large and so
fine-grained that moving a plant a few centimeters might easily put it into a habitat
type for which it is ill adapted. Differences as great as those found between distinct
ecosystems in factors such as temperature, soil aeration, soil moisture, snow cover,
soil fertility, length of the growing season, depth of the thaw, competition, and rates
of herbivory can frequently be traversed in a single step repeatedly across entire
Arctic landscapes. To understand this unique aspect of Arctic plant ecology, it is
necessary to understand how climate and geomorphology interact to produce
unique Arctic vegetation patterns. Understanding the web of feedback interactions
between landscape, moisture, mineral nutrition, and vegetation helps us to understand the complexity of these landscapes and the difficulty of making predictions
regarding climate change in the Arctic.
366
K.M. Peterson
13
367
range of vegetation types at both regional and local scales that frequently reflect
patterns of soil moisture. Different kinds of deserts, sedge grasslands, bogs, and
shrub-dominated communities persist in the Arctic. Names such as polar desert,
polar semidesert, moist tundra, wet tundra, tussock tundra, shrub tundra, coastal
tundra, and bog are commonly used to differentiate between broad classes of
vegetation found within the Arctic.
Polar deserts are unlike other deserts in that the sparseness of vegetation results
from a lack of heat as opposed to a lack of water (Bliss 1997). The growing season
simply does not last long enough for plants to exhaust the soil moisture available
from snowmelt. Density of the vegetation is correlated with differences in growing
season temperature and length, and shifts in vegetation density over decades or
even centuries occur in response to climatic trends. Few plants are capable of
establishment and growth during the extremely short growing season, and seed
production is not always possible. Some communities depend upon exceptional
years for seed production and establishment of new individuals into the population,
while other communities may never set seed and depend entirely upon seeds
dispersed from distant populations or upon vegetative reproduction for recruitment
of new individuals into populations.
Among plants capable of persisting in the polar deserts, Saxifraga oppositifolia
(purple saxifrage) has adapted to a range of habits, with local ecotypes even adapted
to habitats where the short growing season of the High Arctic is further shortened
by late melting snow banks; it does this by increasing metabolic rates and speeding
shoot growth at the expense of accumulating energy and water reserves characteristic of other plants in nearby habitats. Different ecotypes of Saxifraga oppositifolia
also adjust their relative sexual and vegetative reproductive strategies by habitat.
Vegetative reproduction and pseudoviviparity (producing bulbils or plantlets rather
than seeds) are notable adaptations to shortened growing seasons common in Arctic
plant species.
In polar deserts, microhabitats are important to seedling establishment.
Microrelief in soil patterns associated with frost activity can create patterns of
seedling recruitment and survival that persist as vegetation patterning in an otherwise barren landscape (Fig. 1). Centimeters or even millimeters of physical relief
create microclimates useful to plants, with slight variations in solar input, temperature, moisture, and snow cover defining the limits of potential habitats.
At the other extreme of Arctic vegetation, wet landscapes are densely covered
with sedges, grasses, mosses, and forbs (Fig. 2). Growing seasons are longer in the
Low Arctic, and productivity is less limited by heat than by competition for light
and mineral nutrients (Brown et al. 1980). Water from snowmelt frequently remains
ponded on the surface for much or all of the growing season, and emergent aquatic
vegetation and ponds occupy the lowest areas. Soil aeration is poor and root
respiration requires aerenchyma (air passages) in the roots of plants in the wettest
sites. Locally better-drained areas are typically home to woody vegetation, particularly dwarf willows, but many species of dwarf shrubs may be found.
Much lies between the extremes of polar desert and wet coastal tundra, including
mesic or moist tundra, tussock tundra, and shrub tundra. These types are the most
368
K.M. Peterson
13
369
Arctic Flora
The current flora of the Arctic consists of between 2,000 and 2,500 vascular plants
and is recent, being largely a product of the Quaternary Period. The Arctic was
forested during the Tertiary Period, and many of the temperate trees of Eastern
Asia, Eastern North America, and Europe show relationships to one another that
attest to their Arcto-Tertiary origins. With Arctic cooling, these forests retreated
from the Arctic and vegetation similar to that currently in the Arctic began to appear
about three million years ago. Multiple sources are likely for the current flora,
including temperate mountains, especially those of Eurasia, and as much of the
Arctic remained ice-free throughout glacial periods of the Pleistocene, both local
and distant refugia (areas free of ice where plants persisted) complicate past
migratory patterns.
Despite the variability within Arctic vegetation, the flora of the Arctic is highly
conserved, i.e., the same or closely related species comprise similar vegetation
types throughout the Arctic (Polunin 1960). Such circumpolar floristic similarity is
strongest in the Arctic but is also characteristic of the boreal zone (Hulten 1968).
A pattern of increasing floristic affinity with latitude is understood as a consequence, in part, of plant migrations between the old and new worlds (Eurasia and
North America) and the relative youth of the tundra biome.
The area currently occupied by the shallow Bearing Sea between the Russian Far
East and Alaska has been land covered with terrestrial vegetation multiple times in the
past (most recently during the last glacial maximum), allowing plants and animals to
migrate between the continents of Eurasia and North America. This ephemeral
continental connection is called the Bering Land Bridge, and the region including
adjacent lands is often referred to as Beringia. Plan and animal migrations associated
with Beringia help us understand the current high degree of floristic similarity
throughout the Arctic. In addition to such migration, some plants may have established
(or maintained) circum-Arctic affinities through long-distance dispersal. Spores of
mosses are known to travel long distances in the atmosphere, even reaching the jet
stream where they can circle the earth. Animals, especially birds, are effective agents
of seed dispersal, and the Arctic has many migratory birds nesting during summer.
Perhaps unique to the Arctic is long-distance plant dispersal by ice. Plants
growing (and even blooming) have been observed atop glacial ice rafted across
the Arctic Ocean. Icebergs born from Arctic mountains are sometimes discharged
into the Arctic Ocean bearing soils or gravels containing plants or seeds. Massive,
these ice islands sometimes persist for decades locked in Arctic Ocean sea ice as it
circulates in the prevailing clockwise currents (as seen from above the pole). Ice
islands occasionally find foreign shores bearing immigrant plants as passengers.
Arctic salt marshes show an interesting pattern of distribution of the grass
Puccinellia phryganodes with distinct regions having been colonized by plants
with differing numbers of chromosomes, some fertile and some sterile (although
capable of vegetative reproduction by prostrate stems or stolons). Stolons of this
salt marsh grass embedded in sea ice have reportedly been recovered and grown,
providing a potential mechanism for distribution of this widespread Arctic species.
370
K.M. Peterson
13
371
Most substances will warm more quickly than water with an equivalent input of
heat. The amount of heat required to raise the temperature of a substance compared
to the same amount of water is called the heat capacity of the substance. Liquid
water is thus a standard with a heat capacity of one, which is approximately twice
that of either ice or water vapor and more than four times that of dry air. Plants are
primarily composed of water and thus require more heat per unit mass to elevate
their temperature 1 than does an equivalent mass of surrounding air, dry soil, or
dead plant material. Because water has such a high heat capacity, moisture in the
environment is an important control over the thermal behavior of soils, and flowing
water can efficiently transport heat.
Changes of state of a substance also involve gaining or loosing heat. In the
Arctic, water exists in all three states: solid, liquid, and gas. Melting ice or
vaporizing water requires thermal energy input that is not reflected in a change in
the temperature. If heat is added to ice, it will initially warm to thawing temperature
as predicted by the heat capacity of ice, and then it will continue to absorb heat
without a corresponding rise in temperature until all of the ice is melted. Once
melted, continued addition of heat will warm the water, eventually to the point of
vaporization, whereupon the temperature will again remain constant, despite the
continued addition of heat, until all of the water is evaporated. Continued heating at
that point will elevate the temperature of the vapor. As noted above, the rates of
temperature rise in ice, water, and vapor are not equal, being a function of their
distinct heat capacities, but this heat addition results in a change in temperature and
is termed sensible heat. When heat is added without a corresponding change in
temperature, such as when ice is thawing to liquid water or when water is evaporating to vapor (both cases of a change of state), the heat consumed is termed latent
heat. Water has high latent heats associated with its phase transitions compared to
most other substances. Freezing and thawing are an integral aspect of the environment of Arctic plants, and the corresponding latent heat requirement is a significant
energy requirement. High latent heat of fusion for water further implies a correlation between moisture and the thermal characteristics of the environment.
Heat moves by conduction, convection, and radiation or through latent heat
exchange. All are important in understanding the thermal environment of Arctic
plants and plant temperatures. Conduction is the transfer of thermal kinetic energy
between materials in contact with each other. Heat absorbed at the soil surface must
be conducted into the soil. Substances differ in their ability to conduct heat (thermal
conductivity) and since they also differ in their heat capacities, the amount of heat
required to elevate the temperature of soils may differ from one stratum to another.
The rate at which heat moves through the soil is a function of both the thermal
conductivity and the heat capacity and is termed the thermal diffusivity. Not
surprisingly the major variable controlling thermal diffusivity in Arctic soils is
soil moisture content. As heat moves through the soil to the depth of frozen
material, additional heat is required to melt ice in the soil before heat can be
conducted to deeper depths.
Heat is lost to the atmosphere from the soil surface and from plants via convection (as adjacent air is heated and rises away from the surface) and by latent heat
372
K.M. Peterson
loss (as water evaporates at the surfaces). The soil and plant surfaces also exchange
heat via radiation. Heat is absorbed primarily from solar radiation, and heat is lost,
since all objects emit radiation as a function of their temperature. The emitted
radiation is invisible as it is in wavelengths too long to be seen, but thermal imaging
can reveal that warmer objects are brighter in these invisible wavelengths that are
cooler objects. Emission is a function of the temperature of the surface, which along
with evaporation and convection may be influenced by the depth of a layer of
relatively calm air held near the surface called the boundary layer (with the
thickness of this layer largely being a function of the roughness or smoothness of
the surface and the wind speed). Many Arctic plants appear fuzzy or hairy as a
consequence of their morphological adaptations to increase their boundary layer
and reduce heat loss (Fig. 3).
The low temperatures and limited heat in Arctic environments present both direct
and indirect challenges to plants. Plants must geminate, metabolize, grow, and
reproduce at tissue temperatures lower than those of plants in most other ecosystems.
Low tissue temperatures both above- and belowground require plants to adjust
enzymes, adjust membranes, and enhance transport processes to compensate. Altering enzymes to operate at low temperatures (largely through genetic adaptation) is
not always possible, and some morphological and physiological strategies are
missing in the Arctic. For example, the C4 photosynthetic pathway is not found at
all in the Arctic. Some plants compensate for lowered specific enzymatic activity by
increasing the total amount of enzyme (as is common with the photosynthetic
enzyme RuBisCO), but this in turn generates higher demands for nitrogen to build
enzymatic proteins and potentially contributes to nutrient stress. Membrane permeability can be increased at low temperatures by making the lipids more fluid. This is
done by decreasing the degree of hydrogen saturation of the fatty acid tails in the
membrane phospholipids, i.e., increasing the number of double bonds in the hydrocarbon chains in the fatty acid tails of these phospholipids. This is especially
important in roots to allow efficient uptake of mineral nutrients as soils thaw.
13
373
374
K.M. Peterson
13
375
376
K.M. Peterson
temperatures and winter desiccation by snow cover. Taller shrubs are generally
restricted to areas where deep snowdrifts accumulate in winter. Even slightly
elevated areas may be blown nearly free of snow in winter, and plants in such
habitats are generally among those clinging most closely to the soil surface, and it is
here lichens are most abundant.
Production of flowers and ovules and the maturation of seeds may exhibit higher
heat requirements than vegetative growth. As a consequence, some plants rarely
flower and reproduction is primarily vegetative for some species in Arctic habitats.
Others possess adaptations to preserve or augment the heat available for sexual
reproduction. Flower development may be close to the surface where thermal
conditions are best and plants may form cushions, reducing their exposure to
wind. Hairs or fuzz, retarding convective and evaporative heat loss, may protect
buds and flowers (Fig. 3). In some cases flower shapes may be parabolic to concentrate reflected radiation onto the pistil and plants such some members of the genera
Dryas (avens) and Papaver (poppy) combine parabolic flowers with solar tracking to
maximize the heat available in the flower (Fig. 5). This adaptation not only benefits
the developing ovules but also may serve as a thermal reward to pollinators.
Challenges of the Arctic environment are mostly physical, with biological
adaptations less apparent than in species-rich biomes, but two aspects of the Arctic
13
377
Permafrost
Heat is primarily added to the Arctic by solar radiation and is extracted by emitted
thermal radiation from the ground surface back to space. Warm objects emit more
radiation than cold objects, but all objects continually emit thermal radiation. The
balance of incoming and outgoing radiations over time (assuming no other energy
exchanges) determines the temperature of the soil-vegetation surface (and via heat
conduction it also determines the temperatures below the surface of the soil). At the
ground surface the temperature varies daily (and seasonally) as solar input varies,
but at a depth of a few meters, such variations in temperature are very small, being
essentially averaged over time by the process of heat conduction and storage. At a
few meters depth, the substrate temperature remains close to the long-term mean
temperature at the surface. In temperate locations this accounts for the observation
of moderate constant temperature in caves, with caves feeling cooler than outside air
in summer and warmer than outside air in winter. In the Arctic, the mean temperature
is below freezing, resulting in a condition at depth known as permafrost.
Permafrost is not a kind of substance; rather it is a thermal characteristic of a
substance. Any substance that remains frozen (i.e., below 0 C) for a period of
2 years or more is considered to be permafrost. Permafrost may pertain to various
kinds of rock, sand, soil, or ice. Permafrost results from the fact that the average
annual temperature at the surface is below 0 C, and insufficient heat is conducted
into the substrate to raise temperatures at depth above freezing. Most terrestrial
378
K.M. Peterson
landscapes in the Arctic are underlain by permafrost, and because of this lands in
the Arctic are said to be in the zone of continuous permafrost. It is not uncommon
for permafrost to extend deeply into the underlying parent materials, in some cases
as much as hundreds of meters. At some point the interior heat of the earth begins to
have a larger effect than the history of surface temperatures, and this point defines
the bottom of the deep permafrost. In some areas permafrost may be actively
growing deeper due to a recent history of colder climatic conditions, and in other
areas permafrost may be a relic of past climatic conditions and be slowly
disappearing. Of importance to plants is the depth at which the top of the permafrost
is found, or more precisely the depth of the soil column that is thawed at any
particular time, as this is frequently less than the depth to the permafrost. The depth
of thaw reaches a maximum near the end of summer. The perennial maximum
depth of thaw defines the top of the permafrost (as permafrost is defined as
remaining frozen 2 or more years). The zone of thawing soil atop the permafrost
(the active layer) represents the maximum soil volume available to support plant
growth.
Active layer depths vary throughout the Arctic and locally primarily as a
consequence of the ability of various soils to hold and conduct heat. The chief
determinant of soil thermal properties is the amount of water in the soil, with water
increasing both the heat capacity and thermal conductivity of soils. Active layer
depths also vary as a consequence of variations in the amount of solar radiation
absorbed by the surface. Surfaces differ in their reflectivity (albedo) with surfaces
like peat or vegetation absorbing more energy than surfaces like snow or sand
which reflect as much as 90 % or more of the radiation striking them (high albedo).
Ponds and lakes absorb more heat, than vegetated surfaces, contributing to deeper
thaw under water bodies. In some cases this can result in the formation of a talik, or
thaw bulb, which is a pocket of thawed material that persists throughout the winter
under the frozen surface and above the permafrost. Most Arctic landscapes freeze
solid in the winter with the active layer freezing down to the permafrost.
While the bottom portion of the active layer may remain thawed well after the
surface has begun to refreeze at the end of the growing season, at some point in the
fall or winter, the entire soil column becomes frozen and remains frozen until thaw
is initiated at the soil surface at the beginning of the next growing season. Arctic
plants are faced with the challenge of living in a soil medium that freezes
completely during the winter. As the above- and belowground periods of thaw do
not entirely match, Arctic plants have adopted a variety of means to obtain minerals
from soils. Freezing of soils also affects plant habitats indirectly through mechanical and geomorphic processes that create and modify local habitats.
Roots
Arctic plant roots grow in cold soils relative to plants of other biomes. Soil temperatures
are close to air temperatures at the soil surface but decrease to 0 C at the bottom
of the active layer, which throughout most of the Arctic is just a few decimeters.
13
379
Fig. 6 While superficially similar aboveground, these sedges and grass exhibit morphological
differences belowground. On the left, the grass Dupontia fischeri roots are shallow in the warmest
soil near the surface, while the roots of the sedge Carex aquatilis explore deeper and colder soils
and on the right the sedge Eriophorum angustifolium sends its annual root system deepest directly
against the thawing front of the soil (Photo from an experiment by Gaius Shaver)
The depth of thawed soil at the beginning of the growing season is just a few
centimeters at the time of snowmelt, usually in June, and takes a month or more to
reach the full depth of the active layer. Thus, plant roots are generally below 10 C
and some are near 0 C for the entire growing season. At low temperatures diffusion
to root surfaces and the permeability of root membranes are lowered. Low soil
temperatures also affect soil microorganisms, including beneficial mycorrhizae and
rhizosphere organisms. Mycorrhizae are common among Arctic plants, helping to
overcome some of the nutrient limitations imposed by Arctic habitats.
Arctic plants exhibit distinct rooting strategies to cope with the challenges of
cold soils (Fig. 6). Differences in root longevity, mass, and depth are apparent,
even among species that are superficially similar aboveground. At one extreme
exist plants of several species of the genus Eriophorum, including Eriophorum
vaginatum, a widespread dominant plant of the low Arctic, possessing an
annual root system. These plants perennate from the thickened base of their
stems where they store annual reserves accumulated during the growing season,
and both aboveground and belowground structures senesce at the onset of winter. A
new root system and leaves are produced from the stem base at the beginning of the
growing season. The roots of Eriophorum are less dense than those of other
graminoids and relatively short lived compared with roots of other sedges
that may persist many years, but there are two clear benefits of a disposable root
system: First, at the beginning of the growing season, there is no unproductive
investment of root structure frozen into soils, and second, the root tips, which
are the principal point of absorption for the root system, can follow the melting
zone, maximizing root tips in the zone most likely to be richest in nutrients.
380
K.M. Peterson
In contrast, the grass Dupontia fisheri maintains a highly branched fibrous root
system near the soil surface. This strategy concentrates root tips in the warmest and
earliest thawed portion of the active layer. Rooting strategies reflect constraints
other than temperature, and species in bogs or saturated soils frequently exhibit
aerenchyma, internal passages that allow oxygen to diffuse along roots in poorly
aerated soils.
Nutrients
Arctic plants generally experience a limited supply of nutrients, particularly nitrogen and phosphorus ions. A chief cause of low nutrient supply is the limited volume
of soils available to plants. Near-surface permafrost can lead to a situation in which
vegetation experiences a condition similar to pot-bound greenhouse plants, i.e., the
entire soil volume available to plants is already exploited by roots and additional
root growth does not yield additional nutrients. Add to this the fact that low
temperatures potentially place demands on plants for high levels of nutrients, and
it is not surprising that competition for mineral nutrition is a common feature of
tundra vegetation. Low temperatures also retard the decomposition
(or mineralization) of plant materials, and there is little or no input of minerals
from the atmosphere or weathering.
Nutrient availability varies considerably between habitats, and some sites, often
associated with animals, are nutrient rich. Birds and mammals concentrate nutrients, particularly nitrogen around nests, dens, or other areas frequently used. Such
areas are often easily spotted due to the color and luxuriance of the vegetation. In
the mountainous Arctic, areas below cliffs favored by nesting birds, especially sea
birds, are particularly fertile. In lowland tundra even small hummocks stand out in
the flat landscape and are utilized by snowy owls as hunting perches. These are
easily recognized by the vigor and greenness of plants and by the accumulation of
owl pellets (wads of hair and bones of small mammals regurgitated by the owls).
The dens of ground squirrels and Arctic foxes present fertile habitats for plants, and
dens characteristically support plant species with high nutrient requirements such as
Arctagrostis latifolia that is otherwise rare or lacking in the surrounding tundra.
Calcium may be added to soils by antlers. In Caribou, both sexes have antlers that
are annually shed, and calving grounds in particular receive a source of calcium,
which may be an important plant requirement in acidic tundra.
Plants in most habitats are limited by the availability of nitrogen or phosphorus.
Nitrogen is generally available to plants in the form of nitrate or ammonia, with
plants in anaerobic soils more likely to be able to effectively use ammonia. In the
Arctic, it was first discovered that plants can also potentially take up amino acids
directly, and this can contribute significantly to the nitrogen requirements of Arctic
plants. Biological nitrogen fixation (converting atmospheric molecular nitrogen to a
form usable by plants) is accomplished in wet and moist Arctic habitats by
cyanobacteria and lichens, particularly by the lichen Peltigera. Due to the restricted
availability of nutrients, Arctic plants recover large amounts of nutrients from
13
381
senescing tissues. The species Petasites frigidus (Arctic Colts foot) is capable of
recovering in excess of 80 % of the nitrogen and nearly 90 % of the phosphorus
from its senescing leaves.
Inputs of minerals from weathering of rock and mineral soil are low due to low
temperatures. Soil development in general is slowed in the Arctic, and many soils
are quite rocky. Wet tundra soils and bogs are often characterized by peat accumulations that occupy most of the active layer. In sites where organic matter accumulates, permafrost generally aggrades, i.e., the additional accumulation of organic
materials insulates the soil resulting in upward growth of the permafrost. Here,
plants must rely on atmospheric inputs, biological fixation, and minerals released
by decomposition (mineralization). Over time the nutrient economies of such sites
gets tighter, resulting in decreased productivity.
Phosphate ions are subject to leaching and are moved across the landscape by
flowing water at breakup. Low-lying areas are thus generally higher in available
phosphorous that are adjacent soils, helping to account for different species in
adjacent habitats. Windblown sediments from river bars, along with animals, help
move phosphorus back up the hydrological gradient. Mycorrhizae are important to
the nutrient economies of Arctic plants.
382
K.M. Peterson
Geomorphic Processes
Geomorphic processes sculpt local relief. Some processes are similar to those found
outside the Arctic, while others are more typical of the Arctic. No matter what
geomorphic processes are involved, the importance of local microrelief in the
Arctic is paramount to understanding local habitats. Local relief, moisture, and
vegetation are inextricably interlinked.
Annual freezing and thawing of soils leads to several phenomena that collectively create patterns of microrelief. Chief among these is the growth of ice wedges.
Frozen soils and substrates cool and contract during winter. Wherever the matrix of
upper permafrost materials is cemented together by ice, the substrate is inelastic,
and annual cooling produces tension cracks up to 2-m deep. As seen from above,
these cracks form a polygonal pattern similar to those produced in miniature by
13
383
drying mud. Snow, water vapor, and melt water enter these contraction cracks
before summer heat again expands the substrate. Water in its various forms entering
the permafrost is retained as ice and prevents re-expansion of substrate materials
into their former position, creating pressures that deform the substrate and raise
ridges in the overlying soil.
Repeated patterns of cracking lead to annual increments of ice accumulation with
approximately 1 mm added to the width of growing ice formations that are wedgeshaped in vertical cross section (expansion and contraction are greatest near the
surface and diminish to nothing deeper than a couple of meters, leading to the crosssectional wedge shape). On the surface, forces of expansion deform the ground and
elevate soil materials adjacent to growing ice wedges creating a pattern of raised
ridges adjacent to either side of the polygonal ice wedge networks. At the same time
some slumping of materials directly over the growing wedges forms a network of
troughs. Such ice wedge growth leads to very common Arctic patterns of relief
known as low-centered polygons (Fig. 8). These polygons are frequently up to 10 or
more meters across with flat central areas surrounded by raised rims and separated
from adjacent polygons by troughs that mark the location of the underlying ice
wedges. Rims and troughs are frequently less than a meter wide each (but may
exceed 2 m in width) and generally reflect the width of the underlying ice wedge.
Centers of low-centered polygons are poorly drained due to the surrounding rim,
and, as rims continue to grow, polygon centers may become ponds. Polygon troughs
are generally wet but are generally integrated with patterns of regional drainage.
Erosion of low-center polygons typically generates another common landscape
type called high-center polygons. Centers do not change much (if any) in elevation,
but the surrounding troughs deepen, and rims collapse into the deepening troughs,
leaving the polygon centers high and separated from each other only by enlarging
deep troughs. The melting of ice wedges and the collapse of adjacent materials
facilitate such erosion as melt water is drained away. Such features are frequently
found adjacent to streams, lakes, and other areas where lateral drainage is
384
K.M. Peterson
13
385
386
K.M. Peterson
Even in areas of continuous plant cover, contraction cracks may disrupt plant growth,
breaking rhizomes and limiting the vegetative spread of individual plants.
Future Directions
The Arctic has long been considered a fragile ecosystem, subject to disturbance by
development and climate change. Such concerns have led to continuing questions
regarding the susceptibility of Arctic organisms and ecosystems to human
13
387
388
K.M. Peterson
References
Bliss LC. Arctic ecosystems of North America. In: Wielgolaski FE, editor. Ecosystems of the
world 3. Polar and alpine tundra. New York: Elsevier; 1997. p. 551683.
Brown J, Miller PC, Tieszen LL, Bunnell FL. Arctic ecosystem, US/IBP synthesis series, vol. 12.
Stroudsburg: Dowden, Hutchinson and Ross; 1980.
Chapin III FS, Jefferies RL, Reynolds JF, Shaver GR, Svoboda J, Chu EW. Arctic ecosystems in a
changing climate: an ecophysiological perspective. San Diego: Academic; 1991.
Conservation of Arctic Flora and Fauna (CAFF). Arctic flora and fauna: status and conservation.
Helsinki: Edita; 2001. p. 272.
Crawford RMM. Plants at the margin: ecological limits and climate change. Cambridge: Cambridge University Press; 2008. [Hardcover].
Hulten E. Flora of Alaska and neighboring territories: a manual of the vascular plants. Stanford:
Stanford University Press; 1968.
Nuttall M. Encyclopedia of the Arctic. 3rd ed. New York: Routledge; 2004.
Polunin N. Circumpolar Arctic flora. Oxford: Oxford at the Clarendon Press; 1959.
Further Reading
Conservation of Arctic Flora and Fauna (CAFF). Arctic flora and fauna: status and conservation.
Helsinki: Edita; 2001. p. 272. http://www.caff.is/publications/view_document/167-arcticflora-and-fauna-status-and-conservation
Crawford RMM. Plant survival in a warmer Arctic. In: Crawford RMM, editor. Plants at the
margin: ecological limits and climate change. Cambridge: Cambridge University Press; 2008.
Lee JA. Arctic plants: adaptations and environmental change. In: Scholes JD, Barker MG, editors.
Physiological plant ecology (39th symposium of British Ecological Society). Oxford; Malden,
MA, USA: Blackwell Science; 1999.
Pielou EC. A naturalists guide to the Arctic. Chicago: University of Chicago Press; 1995.
Grassland Ecology
14
Contents
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
General Characteristics and Global Distribution of Grasslands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Basic Biology and Ecology of Grasses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Morphology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Population Dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Physiology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Roots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Grasslands, Drought, and Climate Change . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Fire in Grasslands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Grazing in Grasslands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Potential Threats to Grassland Conservation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Grassland Restoration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Future Directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
390
392
398
398
401
402
404
406
408
412
416
418
420
421
Abstract
Grasslands are one of Earths major biomes and the native vegetation of up to
40 % of Earths terrestrial surface. Grasslands occur on every continent
except Antarctica, are ecologically and economically important, and provide
critical ecosystem goods and services at local, regional, and global scales.
Grasslands are surprisingly diverse and difficult to define. Although grasses
and other grasslike plants are the dominant vegetation in all grasslands,
grasslands also include a diverse assemblage of other plant life forms that
contribute to their species richness and diversity. Many grasslands also
support a diverse animal community, including some of the most speciesrich grazing food webs on the planet.
389
390
J. Blair et al.
Introduction
Grasslands and other grass- and graminoid-dominated habitats (e.g., savanna, open
and closed shrubland, and tundra) occur on every continent except Antarctica
(though some grasses do occur there) and occupy about 3040 % of Earths land
surface. They cover more terrestrial area than any other single biome type.
14
Grassland Ecology
391
The extent and diversity of grasslands and related habitats is reflected in their
ecological and economic importance at local, regional, and global scales. For
example, grasslands provide critical habitat for a diverse array of plants and
animals. Grassland soils store tremendous quantities of carbon and other key
nutrients and play a major role in global biogeochemical cycles. There is also a
long and complex relationship between grasslands and humans. Modern humans
are thought to have originated in the open grasslands and savannas of Africa, and
grasslands have provided the template and biological raw material for the development of modern agriculture and associated human societies. The fertile soils that
developed under many grasslands have been plowed and the nutrients mined to
support agricultural production. Domesticated grasses, such as corn, rice, wheat,
oats, and sorghum, have become some of our most important agricultural crops, and
barley was used by Neolithic humans to produce one of the first known alcoholic
drinks. Grasses are not only consumed directly by humans, but they also support the
production of domestic livestock for human use. More recently, several species of
grasses are being widely used or considered as feedstock for biofuel production
(e.g., Panicum virgatum, Miscanthus spp.). It is estimated that as many as 800 million people worldwide rely directly on grasslands for their livelihoods (White
et al. 2000), and virtually everyone uses grassland products (food, fiber, fuel) in
their daily existence. In total, it is clear that grasses and grasslands have played an
important role in the history of humans and will continue to do so in the future.
Grasslands have also played an important role in the development and testing of
ecological theory, such as assessing relationships between species richness and
ecosystem function and as model systems for assessing the impacts of global
changes, including responses to chronic N deposition, elevated CO2 concentrations,
and climate change. This is due, in part, to the relative ease of performing manipulative experiments in grasslands, the sensitivity of grasslands to perturbations, and
the relatively rapid responses they often exhibit to these manipulations. In fact one
of the longest running field experiments in the world is the Park Grass Experiment
at the Rothamsted Experimental Station in England. This experiment was
established in 1856 with the original goal of assessing the effects of various nutrient
amendments on grass yields. The experiment has since been used to address a broad
range of fundamental questions in ecology and evolutionary biology (Silvertown
et al. 2006).
Grasslands also include some of the most endangered ecosystems on the planet,
such as the tallgrass prairies of North America and other temperate grasslands
(Hoekstra et al. 2005). In addition to the historical loss of grasslands to agricultural
expansion, grasslands today are threatened by a broad array of environmental
changes, including climate change, elevated atmospheric carbon dioxide concentrations, increased nitrogen deposition, invasive species, habitat fragmentation,
degradation due to overgrazing, change in natural disturbance regimes (e.g., fire
suppression), and woody plant expansion. Conserving, and in some cases restoring,
these ecosystems will require a solid foundation of ecological knowledge. This
chapter focuses on the ecology of grassland ecosystems and provides the reader
with an introduction to grassland plants and the major abiotic and biotic factors that
392
J. Blair et al.
14
Grassland Ecology
393
Yet this prairie, like most other grasslands, is dominated by just a few species of
grass that comprise the majority of grass cover and contribute the bulk of annual
plant productivity. For example, at Konza Prairie Andropogon gerardii,
Sorghastrum nutans, and Schizachyrium scoparium comprise about 70 % of total
plant cover and up to 90 % of the aboveground net primary productivity (ANPP),
particularly in frequently burned and ungrazed areas. In fact, many grassland types
are described by their dominant species (e.g., bluestem prairie). However, despite
the general prevalence of graminoid plant cover, different types of grasslands are
surprisingly diverse in the richness and cover of non-grass species. Using the Konza
Prairie example, the grasses co-occur with over 400 species of forbs and woody
plants, which provide much of the floristic diversity characteristic of the prairie.
The global distribution of grasslands is extensive, with widespread representation
of grasslands on every continent except Antarctica (Fig. 3). Although grasslands are
394
J. Blair et al.
Fig. 3 Global distribution of grasslands and other ecosystem types dominated by grasses or
graminoid vegetation (Reproduced from White et al. 2000)
14
Grassland Ecology
395
ecosystems, and because wetter forests and woodlands are not as limited by water
availability. These results suggest that grasslands may be especially sensitive to
changes in precipitation amounts or timing in an altered future climate. Seasonality
of precipitation, in addition to total annual amount, is also critical in grasslands. For
example, in North America the area around Washington, DC, is dominated by
eastern deciduous forest, and the annual precipitation is ~102 cm, which is very
similar to the annual precipitation amount (~100 cm) near Lawrence, KS, which is
dominated historically by tallgrass prairie. In spite of similarities in total rainfall
amount, the seasonal distribution of rainfall is very different with over 60 % of the
rainfall occurring in the growing season (April to September) and with drier late
summer months in Lawrence, KS, whereas the precipitation is more evenly distributed throughout the year in Washington, DC. The importance of seasonal patterns
of rainfall in grasslands is apparent in the numerous studies that have used climatic
data and concurrent measurements of ecological processes to identify specific times
of the year (critical climate windows) when precipitation has the greatest effect on
processes such as plant productivity or grass reproductive effort. There are also
significant interactions between rainfall amounts and temperature, and the ratio of
precipitation to the potential evapotranspiration (PET) is often a better predictor of
396
J. Blair et al.
ecological properties and process rates than is mean annual precipitation alone.
Of course, the ability of soils to hold and supply water is also critical, and soil water
dynamics are affected not only by rainfall quantity and intensity but also by
physical characteristics of the soil, such as soil texture and porosity. At local scales,
soil water dynamics in grasslands are often highly correlated with plant physiological processes, plant productivity, and soil microbial activity.
Climatically determined grasslands are those that result from prevailing climatic conditions, as opposed to planted grasslands (pastures or lawns) or those that
represent intermediate successional stages. A characteristic feature of climatically
determined grasslands is that they are subject to periodic droughts, which contributes to the accumulation of highly flammable plant detritus and the occurrence of
periodic fires. Many of the worlds most extensive grasslands occur in the interior
regions of the continents, where annual rainfall amounts are relatively low and
irregularly distributed across the year. Some of these grasslands lie between more
arid deserts and more mesic forests and woodlands, while others occur in the rain
shadows of major mountain ranges. The continental climates of these regions are
often marked by extremes in seasonal temperatures (hot summers and cold winters), to which the plants and animals living there are adapted. For example, at
Konza Prairie in the Central United States, the mean monthly temperature varies
from a January low of 3 C to a July high of 27 C. In temperate grasslands with
such continental climates, a significant proportion of annual rainfall often coincides with the warm growing season, and plant dormancy is a mechanism for
surviving low winter temperatures. Many grassland animals also become dormant
or migrate to avoid harsh winter conditions. In grasslands with a Mediterranean
climate, such as those in the Central Valley of California, dormancy is driven by
summer droughts, and the growing season coincides with seasonal rainfall that
occurs in the relatively warm winter months. Tropical grasslands also exhibit
distinct seasonality based on cyclic annual rainfall patterns, though annual temperatures vary less than in temperate grasslands. Dormancy still occurs, but in this
case it is a response to annual dry seasons that alternate with the rainy growing
season as a result of annual movement of tropical low pressure system boundaries.
Soils of tropical grasslands may also be less fertile than comparable temperate
grassland soils as a result of faster weathering rates under warm year-round
temperatures and soils that are much often much older than in temperate grasslands. Many tropical grasslands also have a greater density of woody shrubs and
trees than do temperate grasslands.
Although many climatically determined grasslands experience seasonal water
deficits and periodic droughts that preclude the establishment of forests in those
regions, some mesic grasslands, such as the tallgrass prairies of North America or
the sourvelds of South Africa, occur in regions where the climate could support
woodland, shrubland, savanna, or even forest vegetation. In these cases, the
persistence of grasslands often depends on recurring disturbances, such as fire
and grazing. Such grasslands may be best thought of as disturbance-dependent
communities, where periodic fires, droughts, and the activities of grazers
are necessary to keep grasslands from transitioning to other ecosystem types.
14
Grassland Ecology
397
In fact, it is generally recognized that climate, fire, and grazing are three key
factors that are responsible for the origin, maintenance, and structure of the most
extensive natural grasslands on Earth. Although the relative importance of fire in
structuring grassland communities tends to be greatest in the most mesic and
productive grasslands, which also burn at more frequent intervals and with greater
fire intensities do to large accumulations of fine fuel in the form of aboveground
grass litter, fires do occur at varying frequencies in most grasslands, including
shortgrass steppe and even desert grasslands. In addition, most grasslands
coevolved with large grazers, and herbivory is an important process affecting
ecological processes at levels ranging from the physiology of individual plants
through population and community dynamics to ecosystem processes and
landscape patterns. Although there are some similarities with respect to the effects
of fire and grazing (i.e., both can be considered disturbances that remove aboveground plant biomass and free up resources), there are importance differences
in their effects on soil resources and plant communities, as well as some important
interactions between fire and grazing in grasslands. The effects of fire and grazing,
and their interactions, are discussed in more detail in later sections of this chapter.
A final characteristic feature of grasslands is a relatively high allocation of plant
biomass belowground (a high root to shoot ratio) and proportionally large inputs of
plant root litter relative to surface litter. Relatively high belowground plant inputs
coupled with relatively slow decomposition rates due to periods of water limitation
can lead to large accumulations of organic matter and nutrients in the soil. In
addition, the limited rainfall characteristic of most grasslands reduces the rate of
weathering and leaching of critical plant nutrients from the rooting zone of grassland soils. The resulting high fertility of grasslands soils is one of the reasons they
have been so widely exploited for agricultural purposes. The accumulation of soil
organic matter is generally positively correlated with water availability, which
stimulates plant productivity more so than decomposition, such that the most
productive grasslands also tend to store the most organic matter and nutrients in
the soil. Although grasslands can occur on a variety of different soil types, the
archetypal dark, rich soils characteristic of many grasslands are known as Mollisols
in the US Soil Taxonomy system or as a Chernozem in the World Reference Base
for Soil Resources. These are the dark, rich soils that formed under the prairie of
North America and the steppes of Europe and that have now largely been cultivated
for use in agricultural production. Grasslands can also occur on other soil types, too.
Many tropical and subtropical grasslands occur on soils that are geologically much
older and therefore more highly weathered than most temperate grassland soils.
These soils may be more depleted in cations and have lower phosphorus availability
than temperate grassland soils. One unique association between soils and grasslands
are the serpentine grasslands. Serpentine soils have a unique chemical composition
due to the type of parent material from which they formed. Serpentine soils
generally have high levels of magnesium and other metals and low concentrations
of calcium. The flora growing on these soils is often very different from surrounding soils growing on more typical soils. In many cases, serpentine grasslands
include species that are uncommon in other habitats.
398
J. Blair et al.
Morphology
The aboveground portion of grasses is organized into tillers individual shoots
growing from the base of the plant. Tillers may be vegetative or reproductive and
consist of one or more repeating units called phytomers, which are the basic
building blocks of all grass shoots. Each phytomer consists of a node and internode
with an axillary bud, cylindrical sheath, and leaf blade (Fig. 5).
Tillers are initiated from undifferentiated cellular tissue (meristematic tissue)
that typically exists just beneath the soil surface. This is an important feature in an
environment that includes periodic disturbances that remove tissues above the soil
surface (i.e., fire and grazing). Additional meristematic tissue in grasses is also
located at the intersections where leaves attach to the tiller (intercalary meristems).
Thus, the oldest portion of a grass leaf is at the tip of the leaf and the top of the plant,
and the youngest portion of a leaf is nearest the stem or the soil surface. For this
reason, when grass blades are eaten, the actively growing plant tissues (intercalary
or basal meristems) are left to produce new growth to replace removed leaf tissue.
The presence of protected meristematic tissue belowground also allows grasses to
survive and regrow when grazed or when fire removes aboveground tissues. This is
an important mechanism giving grasses an advantage in environments with recurring droughts and fires or high grazing pressure (Fig. 6).
An individual grass plant generally consists of multiple joined tillers, but
different grass species show great variation in the way tillers are aggregated as
they expand from their origin. Two general classifications of tiller aggregation
apply to most grasses: bunch-forming (caespitose or tussock) forms that are common in more arid grasslands and sod-forming (rhizomatous) grasses found more
commonly in mesic grasslands (see Fig. 1). Sod-forming grasses utilize stolons
(aboveground stems running along the ground surface) or rhizomes (belowground
stems that occur just beneath the soil surface) to expand laterally through the
14
Grassland Ecology
399
Fig. 5 Structure of the grass plant: (a) General habit (Bromus unioloides); (b) rhizomes; (c)
stolon; (d) rhizome and stolon intergradations (Cynodon dactylon); and (e) the leaf at the junction
of sheath and blade, showing adaxial surface (left) and abaxial surface (right) (Reproduced from
Common Texas Grasses. An Illustrated Guide by F. W. Gould by permission of the Texas A&M
University Press)
asexual production of new tillers (see Fig. 5). Bunch-forming grasses cluster the
production of new tillers around a central stem without rhizome or stolon production. Annual plants and the bamboos are obvious exceptions to these two tiller
classification schemes, as annual plants complete their life history within a single
growing season, and bamboos can produce very large wood-like stems.
Grass leaves are narrow, parallel veined, and characterized by thick-walled cells
that provide rigidity and support that allows them to remain upright despite environmental (i.e., wind) or biotic (trampling) forces. Grasses also have specialized
cells (bulliform cells) that permit leaf rolling during periods of water deficit or
high-light stress, and some species have specialized tissues with air channels
400
J. Blair et al.
14
Grassland Ecology
401
Fig. 7 Scanning electron micrographs of phytoliths. Upper left, Andropogon gerardii; Upper
right, Bouteloua gracilis; Lower left, Festuca sp.; Lower right, Stipa comate (Photos from
E.F. Kelly)
Population Dynamics
Population dynamics of grassland plants are the product of the demography of the
species living there, including life-history traits such as reproductive effort, germination and survivorship, and patterns of mortality. Temperate grasslands can be
divided into two main types based on the life-history characteristics of the dominant
grass species the annual grasslands (i.e., California grasslands) and the perennial
grasslands (i.e., tallgrass prairie). All grasses are flowering plants (Angiosperms)
and nearly all are wind pollinated with a (relatively) simplified floral structure.
Within the annual grasslands, recruitment of new individuals from year to year is
based exclusively on sexual reproduction and germination of seeds by annual (i.e.,
monocarpic) grass species. Seed production and viability are critical parameters
affecting population dynamics, and the soil seed bank is an important reservoir of
new individuals. Annual grass species vary in the longevity of seeds in the soil seed
bank, germination cues, rates of growth, and generation time. In contrast, recruitment of new individuals and population dynamics of perennial grasses are
influenced much less by sexual reproduction and seed dynamics (production,
viability, germination, and growth), but rather are a product of asexual
402
J. Blair et al.
reproduction, and the recruitment of new individuals (really new tillers) is via
clonal stems from existing tillers (Benson and Hartnett 2006). For these perennial
grass species, rhizomes and associated belowground buds are the primary means of
reproduction, and recruitment of individuals from seeds tends to be very low,
except under specific circumstances such as large soil disturbances. Belowground
bud banks in perennial grass species can be very responsive to changing environmental conditions or to disturbances such as fire and grazing, and this may be an
important mechanisms underpinning spatial and temporal variability in the population dynamics and productivity of grasses (Dalgleish and Hartnett 2009).
Physiology
In addition to the morphological adaptations outlined above, grasses possess a suite
of physiological traits that facilitate growth in environments that experience periodic or episodic drought, high light intensity, extremes in temperature, and pulses in
nutrient availability. One of the most fundamental physiological characteristics of
different grass species is the type of photosynthetic pathway used, and this is
another way to distinguish between major grassland types. Throughout the world
today, tropical, subtropical, arid, semiarid, and warm temperate grasslands are
typically dominated by grasses that use a C4 photosynthetic pathway (warm-season
grasses), while grasses using the C3 photosynthetic pathway (cool-season grasses)
are more common in cooler grasslands at higher latitudes or higher elevations.
Most vascular plants (and ~50 % of all grass species) use the C3 photosynthetic
pathway. C3 photosynthesis occurs in leaf mesophyll cells where the enzyme
Rubisco catalyzes a reaction fixing a low-energy carbon source (atmospheric
CO2) to a five-carbon sugar (ribulose bisphosphate), to form two molecules of a
higher energy three-carbon organic acid (3-phosphoglycerate). With energy derived
from the light reactions of photosynthesis, 3-phosphoglycerate is ultimately
reduced to a single six-carbon sugar (glucose) that forms the metabolic template
for all subsequent anabolic pathways in the plant. However, Rubisco is a
nonspecific catalyst and can also catalyze the reaction of O2 with the five-carbon
backbone, ultimately resulting in a net loss of energy to regenerate ribulose
bisphosphate (a process termed photorespiration, which results in a net loss of
fixed carbon). The affinity by Rubisco for O2 over CO2, and therefore photorespiration, increases at higher temperatures and during geologic periods with low
atmospheric CO2 concentrations. These selective pressures are likely to have driven
the evolution of the C4 photosynthetic pathway.
C4 photosynthesis is a more recent physiological and morphological modification
of the C3 pathway, having evolved over 50 different times and in many locations on
Earth (Stromberg 2011). C4 photosynthesis provides a growth rate advantage in the
high-light and high temperature environments typical of many grassland regions
worldwide. In C4 photosynthesis, CO2 is initially captured by the enzyme phosphoenolpyruvate carboxylase (PEP-C) in leaf mesophyll cells to form a four-carbon acid
(oxaloacetate). Oxaloacetate is transported into specialized morphological tissues
14
Grassland Ecology
403
Fig. 8 Grasses with the C4 photosynthetic pathway are more abundant in warmer grasslands of
central US grasslands, whereas C3 grasses show the opposite pattern. Similar patterns occur on
other continents, indicating that differences in biochemical pathways of C fixation play a strong
ecological role in the distribution and success of grasses (From Lauenroth et al. 1999)
named bundle sheath cells that typically surround the leaf conductive tissue. Once in
the bundle sheath, oxaloacetate is decarboxylated, releasing CO2 for Rubisco to fix
and sugars to be formed using the C3 photosynthetic pathway. The primary benefit of
the C4 photosynthetic pathway is the ability to concentrate CO2 within the bundle
sheath essentially eliminating the likelihood of photorespiration and maximizing the
reaction kinetics of carboxylation by Rubisco. As such, the efficiency of energy
capture and conversion into carbohydrates is maximized, and efficient photosynthesis
can be performed in environmental conditions that otherwise would have high
photorespiration (i.e., dry, hot, high-light environments). The advantage of C4 grasses
in warmer climates is evident in the proportions of C4 versus C3 grass species across
latitudinal gradients (Fig. 8).
The C4 photosynthetic pathway has multiple secondary benefits for the grass
species that use this pathway. C4 photosynthesis results in a higher instantaneous
water use efficiency (ratio of CO2 gained to water lost) because PEP-C has a higher
affinity for CO2 than does Rubisco. This allows grasses using the C4 pathway more
flexibility in regulating stomatal openings to reduce water vapor lost from the
leaves via transpiration while maintaining adequate internal CO2 concentrations
for photosynthesis as soils dry down, relative to C3 grasses. The high affinity of
PEP-C for CO2 also allows C4 plants to photosynthesize at higher levels than
404
J. Blair et al.
C3 plants when atmospheric CO2 concentrations are low. As a result, it has been
hypothesized that the C4 photosynthetic pathway may have evolved in response to
declining atmospheric CO2 concentrations during glaciation events of the Earths
history. Finally, because the efficiency of Rubisco is maximized in the high CO2
environment inside the bundle sheath, less total Rubisco is required to maintain a
given rate of carbon assimilation compared to C3 photosynthesis. For this reason,
the photosynthetic nitrogen use efficiency (PNUE) (ratio of C gained per unit N
mass) is higher in C4 plants, allowing for greater productivity in N-limited environments, including many temperate and tropical grasslands.
Roots
As noted previously, most grasslands are characterized by a large investment in root
biomass and a high root to shoot ratio (Fig. 9). However, the root systems of
different grasslands are highly variable in terms of species-specific patterns, total
biomass invested, types of roots produced, and distribution throughout the soil
profile. Many grass species share similar characteristics fine roots that are
highly branched, fibrous in nature, and concentrated in the upper soil profile
(top 2050 cm).
In contrast, the coexisting woody and herbaceous forb species in grasslands have
root types that vary widely in terms of root types (fibrous, taproots, etc.), root depth
distribution, and root to shoot biomass allocation. For this reason, most of our
ability to generalize on the drivers of root structure and function in grasslands has
been focused on the grasses. However, it is important to note that differences in
rooting systems between the grasses and many forbs and woody plants may allow
for differential use of soil resources, such as water and nutrients, and these differences can contribute to coexistence of different life forms in grasslands, as well as
changes in the relative abundance of grasses and other plant life forms under
changing environmental conditions. This concept of niche differentiation among
grasses and woody plants was first described by Heinrich Walter and is known as
Walters two-layer hypothesis (Walter 1971). This hypotheses was originally
intended only for the semiarid savannas of the Southern Hemisphere, but the main
concepts tend to apply to grasslands worldwide; grasses have a relatively fixed
strategy of water uptake focused on surface soils, while woody plants have more
plastic water uptake strategies and typically use considerably more water from
deeper soil depths compared to grasses (Nippert and Knapp 2007).
The amount of root biomass varies markedly among grass species in different
grassland types (mesic semiarid annual grasslands) as well as within a single site
according to interannual variability in climate, topography, soil type, site management (fire and grazing frequency), and by depth in the soil profile. For many
grassland types, the dominant grass species have very high root to shoot ratios
(>3) illustrating a greater allocation of carbon to growth belowground versus
aboveground. While nearly all grasslands are characterized by relatively large
investments in belowground versus aboveground growth, this is typically greatest
14
Grassland Ecology
405
in grasslands with high water or nutrient limitation. In general, dry years (or adverse
environmental conditions) tend to reduce overall grass growth including a reduction
in root production. However, adverse environmental years tend to reduce the
growth of shoots more than the growth of roots in most grasslands, though studies
in the montane grasslands of Yellowstone National Park suggest that roots may be
more sensitive to drought than shoots in some grasslands (Frank 2007). Changes in
root production in response to disturbance tend to be mixed, varying according to
ecosystem type and disturbance legacies. In tallgrass prairies that have been grazed
or recently burned, root production can decrease by ~25 %, as grasses tend to
allocate growth towards new leaf and stem production aboveground. The greatest
reduction in root biomass production in these scenarios is in the uppermost soil
layers (top 10 cm). In some other grasslands, increases in root turnover in the
presence of grazers have been reported.
In addition to high relative belowground biomass (around 7001,000 g m2 in
mesic grasslands), the roots of many grasses extend deep into the soil profile (>2 m
deep in mesic grasslands such as tallgrass prairie). Most grasses do not possess a tap
root, but rather have long fibrous roots that taper with depth. The average depth
distribution of roots in grasslands is generally correlated with mean annual precipitation and the depth distribution of water in the soil profile. Thus, the roots of
grasses in arid grassland are much shallower than those in mesic grasslands
(Fig. 10). Despite the presence of deep roots in some grasslands, the distribution
of root biomass generally declines with soil depth, and majority of the biomass and
total root length is concentrated in the upper soils.
The presence of grass roots at significant depths within the soil led early
grassland ecologists to hypothesize that these roots served as a mechanism for
drought avoidance. This hypothesis presumed that during periods of drought, deep
roots would facilitate water uptake from deep soil zones recharged by infiltration
from winter precipitation and maintain plant growth despite low water availability
in surface soils. A closer examination of the unique physiology and morphology of
406
J. Blair et al.
Fig. 10 Regional gradients in rainfall affects the distributions of major grassland types as well as
mean root depth and root productivity, which in turn affect soil organic matter storage and other
soil properties and processes (From Seastedt 1995)
grass roots has shown that drought tolerance is a more likely strategy used by many
grass species (Nippert et al. 2012). For example, in soils with very low soil
moisture, grasses can maintain carbon uptake despite tremendous negative physical
pressures within the vascular tissues of the roots, stems, and leaves (up to 14 MPa,
or nearly 58 times the pressure of automobile tires!). The ability to withstand these
pressures without collapse is facilitated by vascular tissues with a greater number of
vessels each with a smaller diameter. Thus, while many grasses can be deeply
rooted, the small vessel number and diameter limits the total amount of water that
can be transported from deeper soil depths, compared to the high root biomass and
total root length present in surface soils. The unique physiology, morphology, and
distribution within grassland soils provide a significant advantage for grass roots
compared to forbs and woody plants to tolerate long periods of low water availability during drought.
14
Grassland Ecology
407
Plains region of North America in the early twentieth century. The early 1930s
marked the beginning of a series of successive droughts that resulted in very little
rainfall over much of the Central Plains and extreme reductions in soil moisture in
the top meter of soil. This period, known as the Great Drought, was characterized
by low precipitation (persistent reduction by ~50 % than average), higher wind
speeds, low humidity, and maximum air temperatures that were ca. 56 C above
average maximum values during the summer months (Weaver 1968). The combination of extended severe drought conditions and widespread unsustainable agricultural practices led to the Dust Bowl and the widespread loss of top soil
throughout much of the southern and central Great Plains. Prior to the Great
Drought, Prof. John E. Weaver at the University of Nebraska-Lincoln spent
5 years surveying the community composition of 60,000 sq. miles throughout the
central Great Plains (Weaver and Fitzpatrick 1934). This survey provided the basis
for assessment of changes imposed by the continued drought later in the decade,
and Weaver provided the most detailed assessment of the role of drought on
grassland community structure ever performed.
Initially, the first stages of the drought (19301931) resulted in little change in
grassland community composition (Weaver 1968). However, as the drought continued from 1934 to 1940, it had profound consequences for grassland productivity
and community composition. In the eastern areas dominated by tallgrass prairie, the
initial and most dramatic response to the drought was the desiccation and widespread mortality of the dominant species, primarily big bluestem, Andropogon
gerardii (then classified as Andropogon furcatus); little bluestem, Schizachyrium
scoparium (then classified as Andropogon scoparius); Indian grass, Sorghastrum
nutans; and Kentucky bluegrass, Poa pratensis (Weaver and Albertson 1939). The
loss of cover of the dominant species resulted in the exposure of much bare ground
(estimates range from 36 % to 100 % reductions in basal area of plant cover in the
permanent quadrats studied by Weaver (1968)). The drought eventually impacted
the entire grassland community, with high rates of mortality for forbs, woody
species, and ruderal species. An increase in cover was reported by those species
adapted to drier grasslands to the west (mixed-grass and shortgrass prairie
including western wheatgrass, Agropyron smithii; side-oats grama, Bouteloua
curtipendula; and needlegrass, Stipa spartea). Changes in the relative cover of
species (from tallgrass to shortgrass prairie species) did not occur by immigration of
individuals or seeds, but rather by changes in cover of species that were present, but
less abundant (<1 % of cover), prior to the drought (Weaver and Albertson 1939).
In all, the replacement of true prairie (i.e., tallgrass prairie) by mixed-grass and
shortgrass prairie species occurred over an extensive range (~150 mile wide band)
and within a period of 7 years. While community replacement did occur (from
bluestems to xeric species), large reductions in basal cover (>50 %) persisted. The
dramatic changes recorded during the Great Drought are best expressed by Weaver
(1944, pp. 128129):
The drought has shown clearly that nature has richly endowed True Prairie with many
species, some of which are best adapted to cover the soil, enrich it, and hold it against the
forces of erosion during moist climatic cycles. Others which are then found in such small
408
J. Blair et al.
amounts that they seem almost a non-essential part of grassland rapidly increase to great
abundance and become of great importance when a severe drought cycle occurs. This is
what happened in the 19341940 drought and must have occurred many times in the
historical and geological past, although no written record has been made.
Once the long period of drought ended, bare ground was colonized by ruderal
(i.e., early successional) species common to disturbance (Weaver 1944). Stands of
western wheatgrass, needlegrass, and buffalo grass (Buchloe dactyloides) that had
increased during the drought remained resistant to immediate invasion for the first
few years after drought (although species composition and cover ultimately
returned to pre-drought conditions in the decades to follow). In regions where the
bluestem cover was reduced, but not lost altogether, recovery to pre-drought
abundance occurred within several years via rhizome extension into bare patches.
Finally, for many of the original dominant perennial grasses (bluestems) as well as
the forb species, recovery occurred via dormant rhizomes, root crowns, bulbs, and
corms that persisted in the soil for the duration of the drought (without production
of aboveground stems or leaves). Originally classified as dead years before, these
individuals reinitiated growth 23 years following the drought from their decade of
belowground dormancy [term used by Weaver 1944]. Thus, the recovery of the
tallgrass prairie was spatially and temporally varied with quick recovery (~years)
in locations where species persisted at low abundance but slow recovery (~decades)
in locations where bare patches allowed the development of new grassland communities or replacement by mixed-grass or xeric prairie species.
The responses of grasslands to historic droughts may provide some insights into
possible responses to future climate changes. Many climate change predictions for
regions currently occupied by grasslands include more extreme weather patterns
and increased temperatures, which may combine to reduce soil water availability
and increase plant stress. Past responses to drought suggest that such climate
changes may result in mortality and reduced cover of species adapted to wetter
climates and possible replacement of those species with other adapted to drier
conditions. Such changes in climatic conditions and species distributions would
also be accompanied by changes in a suite of ecological processes, such as primary
productivity, decomposition, nutrient cycling, soil formation, and species interactions. The degree to which species distributions and community boundaries shift in
under a future climate may depend on the rate at which climate changes occur, the
severity of those changes, and whether those changes are transient or represent a
more permanent shift in prevailing climates.
Fire in Grasslands
Grasses produce shoots that when senescent or dormant leave behind fine combustible fuel in the form of surface plant litter (detritus) and standing dead grass
biomass. The accumulation of highly flammable plant litter coupled with periods
of drought, relatively open landscapes, and windy conditions is highly conducive to
large-scale fires (Fig. 11). As a result, fire is (or was) an important force in many
14
Grassland Ecology
409
grasslands around the world, though the frequency and intensity of fire varies as a
function of precipitation (or soil water availability) and aboveground productivity.
Historically, many grassland fires originated as a result of lightning strikes or due to
the activities of aboriginal humans. Once ignited, fire could sweep relatively
unimpeded through large areas of open grassland that lacked natural fire breaks,
and fires are generally thought to have been widespread and common in many of the
extensive grassland regions around the world. The higher productivity of more
mesic grasslands would have promoted more rapid and larger accumulations of
combustible fuel, and so fires were likely more frequent in mesic than arid grasslands. However, even desert grasslands can burn once sufficient fuel accumulates,
and some arid grasslands are more often now as a result of introduced annual
grasses that promote more frequent fires.
The intensity of grassland fires vary, depending on such factors as fuel load
(accumulated biomass), fuel condition (compaction, moisture content, etc.), relative humidity, wind speed, and topography. Grassland fires can be very intense and
can generate sufficient heat aboveground to damage the aboveground shoots of
woody plants (top kill) or even kill entire trees. However, because these fires tend
to move rapidly and much of the fuel is above the ground, most of the heat is
concentrated aboveground and temperatures peak quickly as fire passes. Heat
transfer into the soil is generally small, and soil heating into the range that is
biologically damaging (>60 C) occurs only at the surface. Thus, the belowground
buds and meristematic tissues of the grasses and many other grassland plants are
well protected against even the most intense grass fires. This is an important
contrast to other ecosystems (e.g., forests and woodlands), where the effects of
fire are often associated with an immediate negative impact on plant mortality and
even the effects of soil heating on loss of soil organic matter and nutrients and
changes in soil microbial communities. For grasslands, many of the most significant
effects of fire are indirect and result from changes in the postfire environment,
rather than the effects of the fire per se. Recovery from a fire event in grasslands in
410
J. Blair et al.
terms of new plant growth and accumulated aboveground biomass is generally very
rapid, especially for mesic grasslands. Recovery in more arid or desert grasslands
may take considerably longer.
Changes in natural regimes and/or fire suppression have been implicated as one
of the major drivers of contemporary land-cover change in many grasslands worldwide. In many instances, this is a function of a reduction in the frequency or
intensity of fires relative to their historical occurrence and subsequent increases
in woody plant cover or, in some cases, the conversion of grasslands to shrublands,
woodlands, or forest. However, there are also cases where increasing fire frequency
is the driver of land-cover change, such as the positive feedbacks between grass
cover and fire associated with the spread of invasive fire-prone grasses into ecosystems that were historically less susceptible to fire (e.g., the spread of cheatgrass
(Bromus tectorum) throughout Western US shrublands). Prescribed fire has also
become an important management tool in many grasslands, such as tallgrass
prairies where it is used to limit the growth of woody plants and to promote the
growth and vigor of the dominant C4, or warm-season, grasses. Because of its
importance in the development and persistence of tallgrass prairie, research on the
effects of fire has been a major emphasis of the Konza Prairie Long-Term Ecological Research Program. Fire alters many aspects of prairie ecosystem structure and
functioning. At Konza Prairie, over 20 years of data on the effects of different fire
regimes, including annual spring burning and infrequent burning (every 1020
years), has been amassed. Below examples from these studies have been used to
illustrate some of the ecological effects of grassland fires.
Although fires can occur at anytime of the year, dormant season fires are
generally most common in grasslands. In tallgrass prairie, burning at the end of
winter dormancy (i.e., early spring) is a common management practice. Spring
burning generally increases total plant productivity by stimulating growth of the
warm-season grasses, particularly in times (wet years) or locations (deeper soils)
with adequate soil water available. This is due primarily to the removal of the large
amount of plant detritus (up to 1,000 g m2) that accumulates in the absence of the
fire and the changes in microclimate and soil resource availability induced by the
removal of detritus (Knapp and Seastedt 1986). This detritus acts as a mulch layer,
insulating the soil surface and greatly limiting light availability for emerging plants.
The removal of this accumulated surface detritus and standing dead biomass alters
the energy environment and microclimate of the soil. Direct solar inputs to the soil
increases soil temperatures as much as 20 C in the early spring, relative to
comparable unburned grasslands. The warmer temperatures promote earlier emergence and more rapid spring growth, especially for the dominant warm-season
grasses. In most years, these changes in the soil microclimate promote the growth of
the dominant warm-season grasses, as long as there is adequate water in the soil
profile. However, removal of the detrital layer also enhances evaporation from the
soil surface, and in dry years or shallow soils, this can reduce productivity following
fire. This is also a reason that the effects of fire on plant productivity vary across
precipitation gradients, with positive effects in wetter grasslands and neutral or
negative effects in drier grasslands.
14
Grassland Ecology
411
In tallgrass prairie and other mesic grasslands, the enhanced growth of the
grasses also increases their ability to compete for limiting resources with other
plant species, leading to another effect of frequent fires a reduction in overall plant
species richness and diversity due to reductions in the abundance and cover of many
subordinate species, including the cool-season graminoids and the forbs that provide much of the biodiversity in tallgrass prairie. Thus, frequent burning generally
increases plant productivity, but lowers plant diversity, at least in ungrazed prairie.
The presence of grazers that preferentially graze on warm-season grasses can offset
this effect and changes the relationship between fire and plant diversity, as
discussed in the next section.
In addition to its more apparent effects on prairie vegetation, fire alters nutrient
cycling processes in these grasslands (Blair et al. 1998). The most important effects
involve changes in the cycling of nitrogen. Nitrogen (N) is an essential plant
nutrient which often is in short supply relative to plant demand, and the availability
of N limits plant productivity in many ecosystems. Based on fertilizer studies,
N availability has been shown to limit plant productivity in tallgrass prairies.
However, N limitation is not a universal characteristic of tallgrass prairie and, in
fact, depends on management practices, such as fire and grazing, and on other
external factors, such as climate and topography. In addition to its effects on plant
productivity, N availability can alter competitive interactions among plant species
and, therefore, plant community composition. Nitrogen availability is a major
determinant of plant nutritional quality for herbivores, and the N content of plant
litter influences rates of litter decomposition and therefore the storage of organic
matter in tallgrass prairie soils. Understanding how N cycling processes are altered
by different land-use practices, such as burning, is an important prerequisite to
understanding and predicting grassland ecosystem responses to these practices.
When plant detritus burns, some nutrients are lost with the smoke and gases,
while others are released and deposited in the ash. Much of the nitrogen contained
in surface detritus and plants is volatilized, or converted to gaseous forms, in the
heat of a prairie fire, while other heavier elements such as phosphorus and many
cations are simply deposited in the ash. The volatilization of nitrogen by fire is the
major pathway by nitrogen is lost from the prairie (especially ungrazed prairie), and
frequent fires represent a substantial loss of the prairies nitrogen capital. Nitrogen
cycling in frequently burned prairie is further altered by the responses of the
grasses, which produce more root biomass and produce plant tissue which is
lower in N content, or which has a higher C/N ratio. The increased input of organic
matter with a wider C/N ratio stimulates nitrogen immobilization by soil microbes,
leading to even greater N limitation under frequent burning regimes. Thus, the loss
of N, along with the increased growth of the grasses, greatly reduces the amount of
available N in the soil and increases N limitation for the plants growing in
frequently burned prairie. An important question is how a frequently burned prairie
can maintain higher productivity than unburned prairie, in spite of increased
N limitation. This appears to be, in part, to the increased abundance of warmseason grasses and the high efficiency with which these grasses utilize N, giving
them a competitive advantage over other coexisting plant types.
412
J. Blair et al.
Grazing in Grasslands
Grazing is a form of herbivory in which herbaceous plants (grasses and forbs) are
consumed by herbivores (Fig. 12). This process differs from browsing in which the
leaves and woody twigs are consumed from trees and shrubs. Grazing is, or was
historically, an important process in nearly all grasslands and is considered a key
factor affecting species composition and biomass production in grassland ecosystems. The relationship between grazers and grasslands has developed over millions
of years, and it is likely that grazers and grasslands ecosystems coevolved. Grazers
promote heterogeneity in grasslands by selectively consuming some species while
leaving others, through trampling, soil compaction, soil tunneling, and redistribution of nutrients.
Grazing occurs both aboveground (leaves and stems) and belowground (fine roots
and root hairs) by a wide variety of animal herbivores from microscopic invertebrates
to the large mammalian megafauna. In general, while a relatively low density of the
largest grazers (e.g., bison, wildebeest, zebra) can consume a significant proportion of
plant biomass, many small rodents or numerous invertebrates can have comparable
impacts within the same grassland when their densities are high enough. Grazers can
have a tremendous impact on grasslands through their effects on plant populations and
community composition, on energy flow and nutrient cycling in grassland ecosystems,
and on landscape-level heterogeneity and movement of materials (McNaughton 1985;
Knapp et al. 1999). Although some grasslands (the tallgrass prairies of North America
or the Serengeti grasslands of Africa) appear to be well adapted to relatively high
grazing intensities, other grasslands can be quickly degraded by overgrazing. When
managed in an unsustainable fashion (e.g., overgrazing), large ungulates can significantly impact grassland health and sustainability.
Spatial and temporal patterns of activity by grazers can be greatly affected by
fire and grazing by large herbivores and, in turn, can greatly alter the effects of fire
in grasslands (Fig. 13). These interactive effects of fire and grazing are especially
important in mesic temperate and tropical grasslands. Many large grazers are
attracted to recently burned areas, as the removal of detritus and the emergence
of new grasses provides a high-quality grazing areas. Intensive grazing in these
areas can lead to selection for high-quality grazing tolerant grasses and the formation of a grazing lawn. At the same time, increased grazing intensity in burned
areas removes aboveground biomass that would otherwise accumulate and serve as
fuel for future fires. As a result, fire and grazing in extensive grasslands can be
spatially and temporally dependent on each other and can transform the grassland
landscape into a dynamic mosaic of shifting patches that vary in time since fire,
grazing intensity, and fuel accumulation (Fuhlendorf and Engle 2011). This spatiotemporal interaction of fire and grazing has been referred to pyric herbivory, a
term that highlights the codependence of fire and grazing in many natural grasslands. This same principle is the basis of a proposed alternative management
practice called patch-burn grazing, which is designed to mimic the interaction of
fire and grazers to promote greater heterogeneity and habitat for wildlife in grasslands managed for production of domestic grazers (i.e., cattle).
14
Grassland Ecology
413
414
J. Blair et al.
Thus, grazing can offset the reduction in species diversity that results from frequent
burning of productive grasslands such as tallgrass prairie by reducing grass dominance and increasing plant species diversity in areas that have been burned
(Fig. 14). In xeric grasslands, on the other hand, grazing may lower species
diversity particularly by altering the availability of suitable microsites for forb
species. These effects are strongly dependent on grazing intensity. Overgrazing
may rapidly degrade grasslands to systems dominated by weedy and nonnative
plant species.
Most grazers are highly selective in the plants they consume. This selectivity
results in a landscape with heterogeneous species composition and patchy nutrient
distributions. Plants that lose tissues to grazing must use assimilated carbon and
nutrients to regrow leaves (or roots), leaving less palatable species to grow taller
and increase in number. Many large grazers such as African buffalo, North American bison, or domesticated cattle primarily consume the grasses, allowing less
abundant forb species to increase in abundance and new species to colonize the
space that is made available. In more productive grasslands adapted to the activities
of grazers, grazing can be an important management tool to increase biodiversity
when managed at appropriate stocking rates.
14
Grassland Ecology
415
Grazers also accelerate the conversion of plant nutrients from forms that are
unavailable for plant uptake to forms that can be readily used. Essential plant
nutrients, such as nitrogen, are bound for long periods of time in unavailable
(organic) forms in plant foliage, stems, and roots. These plant parts are slowly
decomposed by microbes, and the nutrients they contain are only gradually released
in plant-available (inorganic) forms. This decomposition process may take several
years. Grazers consume plant tissues, process this material inside the gut, and
excrete nutrients that are available for uptake by plants back onto the landscape.
This nutrient processing happens rapidly compared to the slow decomposition
process, and nutrients are excreted in high concentrations in small patches. Thus,
grazers may increase the availability of potentially limiting nutrients to plants as
well as alter the spatial distribution of these resources.
Some grasses and grassland plants can compensate for aboveground tissue lost
to grazers by growing faster after grazing has occurred. Thus, even though ~50 % of
the grass foliage may be consumed by large grazers, when compared to ungrazed
plants at the end of the season, the grazed grasses may be only slightly smaller, the
same size or even larger than ungrazed plants. This latter phenomenon, called
overcompensation, has not been shown in all grassland ecosystems, but the
ability of grasses to compensate partially or fully for foliage lost to grazers is
well established. Compensation occurs for several reasons including an increase
in light available to growing shoots in grazed areas, greater nutrient availability to
regrowing plants, and increased soil water availability (because less water is being
lost via leaf transpiration compared to an ungrazed dense plant canopy).
As with fire, the impact of grazing on grasslands and the ability of grasslands to
tolerate heavy grazing depend upon where the grassland occurs (usually more
mesic grasslands can recover more quickly than arid grasslands) as well as the
growth form of the grasses within the system: caespitose (bunch-forming grasses)
versus rhizomatous grasses. But another key factor determining the ecological
responses of grasslands to grazing is the evolutionary history of the grassland
(Fig. 15). In general, grasslands with a long evolutionary history of grazers, as in
Africa and North and South America, are very resilient to grazing. The evolution of
this resilience may reflect the migratory nature of most herds of large grazing
mammals. Historically, herds of thousands (and up to millions) of grazers moved
across African and North American landscapes in response to seasonal cues and
availability of resources. While the impact of these large herds has (or had in the
case of North America) a tremendous impact on the grasslands, the animals spend
only a small period of time within a given location, allowing for periods of recovery
before the next grazing event.
Due to the ability of grasses to cope with high rates of herbivory, many former
natural grasslands are now being managed for the production of domestic livestock,
primarily cattle in North and South America and Africa, as well as sheep in Europe,
New Zealand, and other parts of the world. Grasslands present a vast and readily
exploited resource for domestic grazers. However, if not managed properly, grasslands can be easily overexploited with subsequent land degradation, nutrient loss,
and susceptibility to invasion by undesirable plant species.
J. Blair et al.
DIVERSITY
DIVERSITY
long
416
DIVERSITY
short
GRAZING INTENSITY
DIVERSITY
GRAZING INTENSITY
GRAZING INTENSITY
GRAZING INTENSITY
semiarid
subhumid
MOISTURE
14
Grassland Ecology
417
418
J. Blair et al.
Grassland Restoration
Given the ecological importance and extensive loss or degradation of grasslands
globally, it isnt surprising that grassland restoration has become increasingly
important and widespread, especially in locations where substantial areas of native
grasslands have been lost as a result of land-use or land-cover change. Grassland
restoration often takes place on formerly cultivated lands and involves
reintroduction of native species characteristic of grasslands in that particular region.
However, there are other types of grassland restoration, including restorations that
target reductions in woody plant cover in areas that have experienced woody plant
encroachment or those that target the removal of invasive species and their replacement with native grassland species. The motivation for these restoration efforts
varies from restoring native plant biodiversity, to restoring ecosystem processes
that provide environmental benefits (e.g., limiting soil erosion and improving water
quality, sequestering carbon), to providing suitable habitat for regional native
fauna. There are multiple difficulties associated with restoring grassland communities and ecosystems, fragmentation of historically extensive areas of intact grassland, loss of genetic diversity of grassland plant and animal populations, and
insufficient area to include some of the drivers that were historically important in
14
Grassland Ecology
419
shaping grasslands, such of landscape-level patterns of fire and grazing. Nevertheless, there are widespread efforts to restore native grassland diversity and ecosystem functioning.
Much research has focused on restoring temperate grasslands in North America,
particularly in the tallgrass prairie region where the cover of native tallgrass prairie
has declined 8299 % since the 1830s, primarily as a result of cultivation for
agricultural use. Dispersal of native grasslands plants into abandoned agricultural
fields is very limited, and many areas targeted for grassland restoration are isolated
from potential native seed sources. As a result, restoration of these grasslands
typically begins with the introduction of seeds or transplants of native plant species.
One of the earliest attempts to restore tallgrass prairie on ex-arable land began in the
1930s at the Curtis Prairie in Madison, WI. Since then numerous prairie restorations
have been initiated at a range of spatial scales, and recent decades have seen a sharp
increase in efforts to restore prairie for both conservation and research purposes. In
fact, restored grasslands are being used to address a variety of basic and applied
ecological questions, such as the relationship between species diversity and ecosystem function, the role of resource heterogeneity in structuring plant communities, or the role of dominant species in community assembly (Baer et al. 2003,
2005). It has even been suggested that restoration can serve as an acid test of our
understanding of community assembly.
Reestablishing the dominant grass species in restored grasslands is relatively
easy. However, it is difficult to establish and maintain many of the less common
species that provide the majority of biodiversity in native prairies. As a result,
restored grasslands generally have much lower diversity than comparable native
grasslands. Even when initial seed mixtures include a diverse assemblage of
subdominant and rare forbs, establishment of these species may be poor. In addition, the cover of the dominant warm-season grasses tends to increase over time in
many restored grasslands, with a concurrent loss of rarer species, such that diversity
declines over time. Overseeding (adding additional seeds to restored grasslands) is
sometimes used in an effort to overcome potential dispersal limitations and enhance
recruitment of new species in older restorations. However, the underlying reasons
for loss of diversity are unclear, and additional studies are needed to assess the
relative importance of dispersal limitations, interspecific competition, resource
heterogeneity, herbivory, or other factors on limits to diversity in restored
grasslands.
The restoration of grasslands on former agricultural soils can provide other
benefits, including reduced soil erosion, greater nutrient retention, and providing
a sink for atmospheric CO2. One of the well-documented effects of cultivation is the
loss of a significant proportion of carbon stored in the form of soil organic matter.
Cultivation of grasslands reduces inputs of plant-derived new organic matter and
the disruption of soil structure coupled with improved aeration greatly increased
microbial mineralization of stored soil carbon. As a result, grasslands can lose from
20 % to 50 % of their organic carbon content within a few decades of cultivation.
Eventually, these cropland soils come to a new equilibrium soil C content that is
much lower than the grassland soils they replaced. However, if these fields are
420
J. Blair et al.
removed from cultivation and restored with perennial grasses and forbs, the soil
carbon pools will increase as new perennial root systems redevelop, new C inputs
are added to the soil, and soil structure begins to reform. Several studies
have documented significant rates of carbon accrual, generally in the range of
2060 g C m2 year1, and suggested that these rates could persist for decades
until a new equilibrium is reached. It is important to point out, however, that
although some soil C (and N) pools in restored prairie may approach those of native
prairie within a few decades, it may take much longer for other soil properties (e.g.,
soil aggregate structure or soil microbial communities) to recover.
Future Directions
Below are a few suggestions regarding future research directions that are particularly relevant to grassland conservation and management. This is not an exhaustive
list, but rather meant to stimulate further discussions about the scope and directions
of future research required for an improved understanding of grassland ecology and
the maintenance/conservation of these ecosystems around the world.
It is essential to develop a mechanistic understanding of how grasslands are
responding and will respond in the future, to multiple global change phenomena,
including changes such as enhanced N deposition, altered climate, and elevated
CO2 changing land use and land cover. Additional multifactor experiments are
needed to address the interactions of global changes driver that occur in combination. Better forecasting of potential responses to environmental changes will
improve both conservation goals and the sustainable use of grassland resources.
A better understanding of the factors that affect the success of grassland restoration efforts is needed. While many studies have focused on deterministic
factors, such as site preparation, seed sources, and seeding rates, additional
studies that address the relative importance of stochastic factors (e.g., climatic
variability, in establishment years) are also needed. This information will be
critical for designing more effective methods of restoring grassland in areas
where they have been degraded or extirpated.
Effective management and conservation of grasslands will require a better
understanding of social and economic drivers. One example of a newly emerging
threat is the increase in restrictions on the use of grassland fires for management
and conservation due to human health concerns. There is a need to explore other
methods to minimize the negative effects of burning (e.g., impacts of smoke on
air quality) in areas where fire is essential for maintaining grassland flora and
fauna or perhaps ways to simulate some of the major ecological effects of fire
to achieve desired management goals.
Understanding the abiotic and biotic conditions that result in variable responses
to grazing in different grasslands has both basic and applied significance. Many
studies report contrasting effects to grazing, for example, with respect to root
productivity and belowground carbon allocation. Similar conflicting results have
been reported to for a suite of other responses. The occurrence of grazing in most
14
Grassland Ecology
421
References
Anderson RC. The historic role of fire in the North American grassland. In: Collins SL, Wallace
LL, editors. Fire in North American tallgrass prairies. Norman: University of Oklahoma Press;
1990.
Archibold OW. Ecology of world vegetation. London/New York: Chapman and Hall; 1995.
Baer SG, Blair JM, Knapp AK, Collins SL. Soil resources regulate productivity and diversity in
newly established tallgrass prairie. Ecology. 2003;84:72435.
Baer SG, Collins SL, Blair JM, Fiedler A, Knapp AK. Soil heterogeneity effects on tallgrass prairie
community heterogeneity: an application of ecological theory to restoration ecology. Restor
Ecol. 2005;13:41324.
Benson E, Hartnett DC. The role of seed and vegetative reproduction in plant recruitment and
demography in tallgrass prairie. Plant Ecol. 2006;187:16377.
Blair JM, Seastedt TR, Rice CW, Ramundo RA. Terrestrial nutrient cycling in tallgrass prairie. In:
Knapp AK, Briggs JM, Hartnett DC, Collins SL, editors. Grassland dynamics: long-term
ecological research in tallgrass prairie. New York: Oxford University Press; 1998.
Blecker SW, McCulley RL, Chadwick OA, Kelly EF. Biologic cycling of silica across a grassland
bioclimosequence. Global Biogeochem Cycles. 2006;20, GB3023. doi:10.1029/
2006GB002690.
Briggs JM, Knapp AK, Blair JM, Heisler JL, Hoch GA, Lett MS, McCarron JK. An ecosystem in
transition: causes and consequences of the conversion of mesic grassland to shrubland.
BioScience. 2005;55:24354.
Collins SL, Knapp AK, Briggs JM, Blair JM, Steinauer E. Modulation of diversity by grazing and
mowing in native tallgrass prairie. Science. 1998;280:7457.
Dalgleish HJ, Hartnett DC. The effects of fire frequency and grazing on tallgrass prairie productivity and plant composition are mediated through bud bank demography. Plant Ecol.
2009;201:41120.
Fargione JE, Hill J, Tilman D, Polasky S, Hawthorne P. Land clearing and the biofuel carbon debt.
Science. 2008;319:12358.
Frank DA. Drought effects on above and below ground production of a grazed temperate grassland
ecosystem. Oecologia. 2007;152:1319.
Fuhlendorf SD, Engle DM. Restoring heterogeneity on rangelands: ecosystem management based
on evolutionary grazing patterns. BioScience. 2011;51:62532.
Hoekstra JM, Boucher TM, Ricketts TH, Roberts C. Confronting a biome crisis: global disparities
of habitat loss and protection. Ecol Lett. 2005;8:239.
Knapp AK, Seastedt TR. Detritus accumulation limits productivity of tallgrass prairie. BioScience.
1986;36:6628.
Knapp AK, Blair JM, Briggs JM, Collins SL, Hartnett DC, Johnson LC, Towne EG. The keystone
role of bison in North American tallgrass prairie. BioScience. 1999;49:3950.
422
J. Blair et al.
Lauenroth WK, Burke IC, Gutmann MP. The structure and function of ecosystems in the central
North American grassland region. Great Plains Research 1999;9:22359.
McNaughton SJ. Ecology of a grazing ecosystem: the Serengeti. Ecol Monogr. 1985;55:25994.
Milchunas DG, Sala OE, Lauenroth WK. A generalized model of the effects of grazing by large
herbivores on grassland community structure. Am Nat. 1988;132:87106.
Nippert JB, Knapp AK. Linking water uptake with rooting patterns in grassland species.
Oecologia. 2007;153:26172.
Nippert JB, Knapp AK, Briggs JM. Intra-annual rainfall variability and grassland productivity: can
the past predict the future. Plant Ecol. 2006;184:6574.
Nippert JB, Wieme RA, Ocheltree TW, Craine JM. Root characteristics of C4 grasses limit
reliance on deep soil water in tallgrass prairie. Plant and Soil. 2012;355:38594.
Prasad V, Stromberg CA, Alimohammadian H, Sahni A. Dinosaur coprolites and the early
evolution of grasses and grazers. Science. 2005;310:117790.
Seastedt TR. Soil systems and nutrient cycles of the North American prairie. In: Joern A, Keeler
KK, editors. The changing prairie. New York: Oxford University Press; 1995.
Silvertown J, Poulton P, Johnston E, Edwards G, Heard M, Biss PM. The park grass experiment
1856-2006: its contribution to ecology. J Ecol. 2006;94:80114.
Stromberg CAE. Evolution of grasses and grassland ecosystems. Annu Rev Earth Planet Sci.
2011;39:51744.
Walter H. Ecology of tropical and subtropical vegetation. Edinburgh: Oliver & Boyd; 1971.
Weaver JE. Recovery of midwestern prairies from drought. Proc Am Philos Soc. 1944;88:12531.
Weaver JE. Prairie plants and their environment: a fifty-year study in the Midwest. Lincoln:
University of Nebraska Press; 1968.
Weaver JE, Albertson FW. Major changes in grassland as a result of continued drought. Bot Gaz.
1939;100:57691.
Weaver JE, Fitzpatrick TJ. The prairie. Ecol Monogr. 1934;4:109295.
White R, Murray S, Rohweder M. Pilot analysis of global ecosystems (PAGE): grassland ecosystems. Washington, DC: World Resources Institute (WRI); 2000.
Further Reading
Axelrod DI. Rise of the grassland biome, Central North America. Bot Rev. 1985;51:163201.
Beerling DJ, Osborne CP. The origin of the savanna biome. Glob Chang Biol. 2006;12:202331.
Borchert JR. The climate of the central North American grassland. Ann Assoc Am Geogr.
1950;40:139.
Collins SL, Wallace LL, editors. Fire in North American tallgrass prairies. Norman: University of
Oklahoma Press; 1990.
French N, editor. Perspectives in grassland ecology. Results and applications of the United States
international biosphere programme grassland biome study. New York: Springer; 1979.
Gibson DJ. Grasses and grassland ecology. New York: Oxford University Press; 2009.
Havstad KM, editor. Structure and function of a Chihuahuan desert ecosystem: the jornada basin
long-term ecological research site. New York: Oxford University Press; 2006.
Knapp AK, Briggs JM, Hartnett DC, Collins SL, editors. Grassland dynamics: long-term ecological research in tallgrass prairie. New York: Oxford University Press; 1998.
Lauenroth WK, Burke IC, editors. Ecology of the shortgrass steppe: a long-term perspective.
New York: Oxford University Press; 2008.
McClaran MP, Van Devender TR. The desert grassland. Tucson: University of Arizona Press;
1997.
14
Grassland Ecology
423
Oesterheld M, Loreti J, Semmartin M, Paruelo JM. Grazing, fire, and climate effects on primary
productivity of grasslands and savannas. Ecosyst World ISSU. 1999;16:287306.
Osborne CP. Atmosphere, ecology and evolution: what drove the Miocene expansion of the C4
grasslands? J Ecol. 2008;96:3545.
Risser PG, Birney EC, Blocker HD, May SW, Parton WJ, Weins JA. The true prairie ecosystem.
Stroudsburg: Hutchinson Ross; 1981.
Sala OE, Parton WJ, Joyce LA, Lauenroth WK. Primary production of the central grassland region
of the United States. Ecology. 1988;69:405.
Samson F, Knopf F. Prairie conservation in North America. BioScience. 1994;44:41821.
Weaver JE. North American prairie. Lincoln: Johnsen Publishing; 1954.
15
Anna R. Armitage
Contents
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
History . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Stressors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Salt Marshes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Zonation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Case Study: Plant-Animal Facilitation in a New England Salt Marsh . . . . . . . . . . . . . . . . . . . .
Mangroves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Mangrove Stress Adaptations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Zonation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Case Study: Plant-Animal Interactions on Mangrove Islands in Florida . . . . . . . . . . . . . . . . . .
Future Directions: The Salt Marsh-Mangrove Ecotone: A Developing Field . . . . . . . . . . . . . . . .
Ecosystem Functions and Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Water Quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Nutrient Cycling and Storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Erosion Control and Surge Buffer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Nursery Habitat . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Recreation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Management Issues and Strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Sea Level Rise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Freshwater Diversion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Eutrophication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Policy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Restoration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Future Directions: Integrating Science and Restoration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
426
427
427
431
431
433
437
437
440
443
444
445
447
447
448
449
449
450
450
451
452
452
452
453
454
454
425
426
A.R. Armitage
Abstract
Coastal wetlands are plant communities at the land-sea interface. Two common types of coastal wetlands are salt marshes and mangrove swamps.
Marshes are dominated by nonwoody grasses and shrubs; mangrove swamps
are dominated by trees.
The global distribution of salt marshes and mangroves is governed by temperature: most mangrove species cannot tolerate freezing temperatures, so
they grow in warmer tropical and subtropical latitudes. Marshes are more
common in cooler temperate latitudes.
Salt marshes and mangroves overlap in some subtropical regions; these areas
may experience shifts in species composition in response to climate change.
The dynamics and ecological consequences of these shifts are important
topics for future research.
Plants in coastal wetlands are adapted for abiotic stressors including
prolonged inundation, which causes soil anoxia, and high salinity.
Salt marshes exhibit predictable zonation patterns, where the distribution of
species within a site varies with small changes in elevation. These zonation
patterns are driven by species-specific adaptations to abiotic stressors and by
interspecific competition. Zonation patterns in mangrove swamps are more
variable.
Coastal wetlands provide a variety of ecosystem services to human communities: wetlands can improve water quality, store nutrients, and buffer against
erosion and storm surge and provide nursery habitat for commercially and
recreationally important fishery species.
Current management issues in coastal wetlands include encroaching
suburban and agricultural development, sea level rise, nutrient enrichment
and eutrophication from agricultural runoff and treated sewage discharge, and
freshwater diversion.
The policies regulating development on coastal wetlands are complex and
dynamic. Restoration is the most common approach to mitigate for anthropogenic impacts. An understanding of wetland ecology is crucial to making
wise decisions concerning the nature and direction of restoration projects.
Introduction
Coastal plant communities are broadly defined as those habitats shaped by terrestrial and marine influences. Many, though not all, coastal habitats can be defined as
wetlands; the ecology and management of those habitats are covered in this
chapter. Wetlands are defined by the United States Army Corps of Engineers by
the presence of three features: (1) wetland hydrology, inundation or saturation for
at least part of the growing season; (2) hydric soils, soils that are anoxic (containing
little or no oxygen for at least part of the growing season; this condition usually
develops when soils are inundated with water); and (3) hydrophytic vegetation,
vegetation adapted to wet conditions.
15
427
The coastal wetlands covered in this chapter are often located within estuaries.
An estuary is a semi-enclosed body of water where freshwater from rivers or
streams mixes with oceanic waters, creating brackish (slightly salty) conditions.
Tidal movement and riverine freshwater input are variable, causing spatial and
temporal variations in salinity (Fig. 1). Freshwater input supplies estuaries with
sediment, organic matter, and critical nutrients such as nitrogen, phosphorus, and
iron. Tidal marine input brings in animal larvae and other essential nutrients such as
sulfate and bicarbonate. The combination of these freshwater and marine inputs
makes estuaries highly productive habitats.
History
Many early human cultures lived in harmony with wetlands, using these productive
habitats to obtain food, fuel, and shelter. However, beginning in the 1700s, and
perhaps even earlier, many agriculture-based cultures viewed wetlands as fallow
areas with no cultivation value and as breeding grounds for disease-carrying
insects. For decades, wetlands were drained for agriculture or cleared and filled
for development. By the mid-twentieth century, the resultant wetland losses totaled
more than 50 % worldwide; up to 80 % of that loss may be attributable to
agricultural expansion (Dahl 1990).
By the 1970s, however, the links between wetland habitats and vital coastal
ecosystem services fishery support, erosion control, water quality improvement
had become better understood. The rate of development slowed, impacts became
better managed, and restoration began in earnest. Now, the need to protect and
manage these habitats has emerged as a top priority in coastal management. These
ecosystem services and management and restoration challenges will be discussed in
more detail later in the chapter.
Stressors
Freshwater and marine inputs can augment estuarine productivity, but those inputs
also create abiotically stressful conditions. Plant communities are particularly
strongly influenced by salinity and flooding, which is usually accompanied by
anoxia (no oxygen) or hypoxia (low oxygen) in the soil.
Most plants in coastal wetlands are halophytes tolerant of high salt levels.
Halophytes can withstand some amount of salt in their tissues, but even the most
halophytic species must be able to avoid excessive salt accumulation. High concentrations of salt ions can have many negative impacts on plants: salt ions can be
toxic, create an osmotic imbalance that prevents uptake of water even when
inundated, and repel and prevent uptake of positively charged nutrients like NH4+
(ammonium). At the ecosystem level, saline coastal wetlands often have lower
plant biomass but faster rates of decomposition, which in turn yields slower rates of
Fig. 1 Left: typical salinity gradient in the Galveston Bay estuary (Texas, USA), depicting lower salinity (light blue shades) near the riverine inputs and
higher salinity (dark blue shades) near the marine input. Right: salinity gradient in the bay during an exceptional drought in 2011 (Data provided by the
Galveston Bay Estuary Program)
428
A.R. Armitage
15
429
Fig. 2 Simplified conceptual model depicting the relationships between plant productivity,
salinity, oxygen availability, and nitrogen and sulfur cycling in wetland sediments. The gray
cloud represents the relative size of the oxygenated rhizosphere. Solid arrows represent active
processes; dashed arrows represent inhibited or reduced processes
430
A.R. Armitage
Fig. 3 Rhizome cross sections of two wetland plants, showing the hollow spaces forming the
aerenchyma tissue. (a) Spartina alterniflora, a low-elevation grass species with extensive aerenchyma. (b) Spartina patens, a mid- to high-elevation grass species with less aerenchyma tissue
(Photo credit A.R. Armitage)
extend from the leaves to the roots, providing a low-resistance internal pathway for
the transport of oxygen from the leaves above the water to the submerged tissue
(Fig. 3). Aerenchyma forms from the collapse of cortex cells in programmed cell
death (apoptosis). Through aerenchyma, oxygen is transported to the roots to be
used for metabolic processes. The subsequent oxygenation of the rhizosphere
(zone surrounding the roots of plants) can lower sulfide production and reduce
sulfide toxicity. If, however, sulfide production is extremely high, aerenchyma can
become occluded by callus tissue (cells that grow over wounds), leading to plant
dieback events.
Both salinity and low soil oxygen levels can potentially impact nitrogen cycling
in coastal wetlands, largely because some steps in the nitrogen cycle are oxygen
dependent, and others require anoxic conditions. The simplified conceptual diagram in Fig. 2 illustrates some of the key interactions among salinity, oxygen levels,
and the nitrogen cycle. High salinity is linked to lower primary productivity, thus
lowering oxygen production and transport to the rhizosphere. Lower oxygen levels
in the rhizosphere facilitate the anaerobic reduction of sulfate to hydrogen sulfide.
Hydrogen sulfide (H2S) is toxic at high concentrations, which further reduces
productivity and creates a feedback that maintains anoxic conditions. Nitrogen
fixation, the conversion of atmospheric nitrogen (N2) to ammonium (NH4+), is an
anaerobic process that occurs at a relatively rapid rate in most anoxic wetland soils.
However, H2S blocks ammonium uptake, further reducing productivity and contributing to the feedback loop that maintains anoxic soil conditions. Denitrification
is also lower at high salinity, in part due to the salt-mediated inhibition of nitrification, an aerobic (oxygen dependent) process that converts ammonium into
nitrites and then nitrates (Fig. 2).
Different types of wetlands are typically defined by the character of their plant
communities. Swamps are wetlands dominated by trees or shrubs; marshes are
primarily composed of herbaceous, nonwoody vegetation such as grasses, rushes,
15
431
sedges, and forbs. Both swamps and marshes can occur in marine and freshwater
habitats; this chapter will focus on two common types of coastal marine wetland
communities: salt marshes and mangrove swamps.
Salt Marshes
Salt marshes are defined as those marshes subjected to regular tidal flooding by salt
water. Salt marshes occur in estuaries and along marine coastlines, primarily in
temperate latitudes. In tropical regions, the short-stature grasses and forbs in salt
marshes are generally outcompeted by the taller vegetation in mangrove forests,
which will be addressed later in this chapter. A typical salt marsh can be subdivided
into several zones based on elevation relative to sea level (Fig. 4). Each of these
zones varies in salt and flooding stress; the plants in each of these zones are adapted
to those conditions.
Zonation
The border zone between salt marshes and nontidal upland habitat is characterized
by plants that can grow in moderately saline soils but are intolerant of flooding.
Plants in this marsh border zone along the high tide line often lack aerenchyma
tissue, making them sensitive to flooding and associated soil anoxia. For example,
the marsh elder, Iva frutescens, a typical marsh border plant in the Gulf of Mexico,
experiences reduced growth and higher mortality if the roots are inundated for as
little as 8 % of the growing season (Fig. 5; Thursby and Abdelrhman 2004).
Below the marsh border zone is a large zone broadly often referred to as high
marsh. This zone covers a relatively wide elevation range that encompasses a
variety of flooding regimes. In this zone, salts tend to accumulate in the soils due
to regular but brief tidal flooding followed by evaporation, especially in the more
seaward region of the zone. Soil salinities can be more than double that of ambient
floodwater. Despite this stressor, plant diversity tends to be high relative to lower
elevations (Fig. 6), in part because there are many different adaptations to salt
stress. Few plants in this zone are tolerant of prolonged flooding many have
reduced or absent aerenchyma (Fig. 3b).
The lowest vegetated elevation zone in a salt marsh is the low marsh. Soil salinity
is close to that of ambient floodwater. Plant species in this zone must be able to
produce extensive aerenchyma in order to withstand prolonged flooding (Fig. 3a).
Few plant species can survive the anoxic conditions associated with extensive
flooding, so the low marsh zone has relatively low plant diversity. On the east and
Gulf coasts of the United States, the low marsh zone is dominated by Spartina
alterniflora (Fig. 4). This grass species occurs in all tidally flooded zones of salt
marshes, but it grows taller at lower elevations than at higher elevations (Fig. 7). The
mechanisms driving this morphological variation are complex; genetic differences
and environmental influences both contribute to tall- and short-form morphology.
432
A.R. Armitage
Fig. 4 Zonation patterns in a salt marsh. In this picture, the marsh border zone is dominated by the
marsh elder, Iva frutescens. The high marsh zone is comprised of grasses such as Spartina patens
(marsh hay; lighter green) and the rush Juncus roemerianus (black rush; darker green). The low
marsh zone is dominated by the grass Spartina alterniflora (smooth cordgrass) (Photo credit
A.R. Armitage)
Fig. 5 Excerpt from Fig. 7 in Thursby and Abdelrhman (2004). Relationship between mean stem
diameter for older stems of Iva frutescens and the duration of flooding (as percent of growing
season) at the root zone (10 cm below soil surface). Percent flooding values are based on elevation
measurements made near the same location that the stem samples were taken. Vertical bars are 2
SE. The means are of 10 stems except for Fox Hill Cove (FOX) and Jenny Creek (JEN) (n 30)
and Mary Donavon Marsh-1 (DON1) (n 20); ( p < 0.01) (Reprinted with permission from
Springer-Verlag)
Within marsh zones, a microhabitat called a salt pan can form. Salt pans are
unvegetated or sparsely vegetated patches, usually in the high marsh, that are
characterized by very saline soil. There are several mechanisms for the formation
of salt pans (Boston 1983). For example, wrack (floating organic debris) deposition
15
433
Fig. 6 Simplified conceptual model depicting the relative importance of abiotic stressors and
biotic interactions at different elevations within salt marshes. The predominant factor in each
elevation zone is highlighted in the boxes at the top of the graph
can cover underlying vegetation (Fig. 8a). When it is eventually washed out following a high spring tide (during full or new moon phases), the ground underneath will
be devoid of vegetation. Alternative mechanisms of salt pan formation include ice
scouring, which can remove large clumps of marsh vegetation in the winter, or
waterlogging in small topographic depressions, which can cause mortality of
established plants. In all cases, after initial formation of the bare patch, evaporation
will rapidly raise soil salinity, often to more than twice as high as ambient seawater.
High salinity will depress seed germination and inhibit plant invasion, preventing
recolonization and maintaining the salt pan microhabitat for long periods of time.
Vegetation in salt pans is typically restricted to a few individuals of extremely salttolerant species (e.g., Sarcocornia spp.) and blue-green algae (cyanobacteria)
(Fig. 8b). Although these microhabitats have little vegetation, they provide important roosting habitat for many coastal bird species (Fig. 8c).
434
A.R. Armitage
Fig. 7 Tall and short forms of Spartina alterniflora (Photo credit A.R. Armitage)
elevations, where it co-occurs with several marsh fauna species. The marsh grasses
and fauna have a close facultative mutualistic relationship, where each benefits
from the other, though they do not completely rely on each other for survival. One
common faunal group in salt marshes is comprised of fiddler crabs (Uca spp.),
which excavate extensive burrows. In a set of experiments, Bertness (1985)
removed crabs from high-density, low-elevation zones and added crabs to
low-density, high-elevation zones. These experiments revealed several mutualistic
interactions between crabs and smooth cordgrass. Crab burrowing activity oxygenates the sediment, augments drainage, and increases the decomposition of organic
matter, all of which increase smooth cordgrass above- and belowground productivity. Crabs benefit from this association as well smooth cordgrass roots substantially increase the integrity of crab burrows. This positive feedback between
smooth cordgrass and fiddler crabs is strongest within the low marsh elevation, just
above the marsh vegetation-water interface. Burrows excavated at the marsh edge,
in softer, wetter sediment with few roots, will rapidly collapse. At high marsh
elevations, denser root mats interfere with the ability of fiddler crabs to excavate
burrows. Therefore, the strength of the fiddler crab-smooth cordgrass facilitation is
greatest at the upper edge of the low marsh, where there is a maximized mutual
benefit for plants (anoxia stress is alleviated) and crabs (burrow integrity is
increased).
15
435
Fig. 8 (a) Wrack deposition in the high marsh zone of a salt marsh. Wrack has accumulated
between stands of Borrichia frutescens (sea oxeye daisy, with yellow flowers) and short-form
Spartina alterniflora. Previously covered patches that have turned into salt pans are visible in the
background. (b) Fully formed salt pan with sparse succulent vegetation and cyanobacterial mats
(visible as blackened patches on the soil). (c) Black skimmers (Rynchops niger) roosting in a salt
pan (Photo credit A.R. Armitage)
Another common animal in New England salt marshes is the ribbed mussel
(Geukensia demissa). These bivalves require a surface for attaching anchoring
filaments, and smooth cordgrass stems and roots provide a suitable substrate
(Bertness 1992). Mussels can be particularly dense along the seaward edge of the
tall smooth cordgrass zone. The anchoring filaments bind smooth cordgrass stems
together, which in turn increases sediment stabilization and decreases erosion.
Mussels deposit waste products that provide nutrients for plant growth (Jordan
and Valiela 1982), resulting in increased aboveground and belowground productivity (Fig. 9; Bertness 1984). Mussels also benefit from this association mussel
growth and survivorship is higher for mussels in smooth cordgrass beds (Stiven
and Kuenzler 1979). Smooth cordgrass benefit mussels by providing an attachment
substrate and may also supply organic matter as an indirect food source
(Bertness 1984).
436
A.R. Armitage
Fig. 9 Excerpt from Fig. 3 in Bertness (1984). Summary of aboveground Spartina alterniflora
parameters in mussel manipulation experiments done on the marsh edge during the 1981 and 1982
growing seasons. Control quadrats; mussel removal quadrats (SE) (All data are for 0.25-m2
quadrats). *P < .05, ANOVA in comparison to control within years. **P < .01, ANOVA in
comparison to control within years (Reprinted with permission from the Ecological Society of
America)
15
437
Mangroves
Mangrove swamps are dominated by halophytic (salt tolerant) trees that live at the
land-sea interface. In an example of convergent evolution, mangrove species
evolved from non-mangrove plant lineages independently many different times.
In fact, mangroves occur in over 30 families of dicots (class Magnoliopsida).
Therefore, trees that are called mangroves are not necessarily closely related in
an evolutionary sense. Mangrove species differ in their stress adaptations and in
their degree of stress tolerance. However, most mangroves are intolerant of freezing
temperatures, which limits their distribution to tropical and subtropical latitudes
(Fig. 10; Giri et al. 2011). There are over 65 species of mangroves worldwide, with
the highest diversity in the Indo-Pacific and Indian Oceans; about four species occur
in North America and the Caribbean.
Fig. 10 Excerpt from Fig. 1 in Giri et al. (2011). Mangrove forest distributions of the world 2000 (Reprinted with permission from Blackwell Publishing
Ltd)
438
A.R. Armitage
15
439
Fig. 11 Excerpt from Fig. 4 in Pi et al. (2009). Cross sections of root tip, basal zone (4 cm from
the root tip), and mature zone (8 cm from the root tip) of Excoecaria agallocha, Lumnitzera
racemosa, and Bruguiera gymnorrhiza (cross sections with thickness of 10 m were made and
photographed, scale bars equal to 200 m; E+H epidermis and hypodermis, Ar aerenchyma air
spaces, Ct cortex, SW suberized walls) (Reprinted with permission from Elsevier BV)
by facilitating rapid establishment soon after dispersal. In mangroves, vivipary protects new, vulnerable seeds from salt water stress, allows nutrient uptake from the
parent plant under low salt stress, and reduces chloride inhibition of germination.
Propagules can float after being released from the parent tree, facilitating longdistance dispersal. Rooting is initiated when favorable habitat is encountered.
A striking morphological characteristic of many mangroves is their complex
aerial root structures, which primarily function as adaptations to flooded conditions.
440
A.R. Armitage
Fig. 12 (a) Propagules of the red mangrove, Rhizophora mangle, still attached to the parent tree.
(b) Rooted propagule of the black mangrove, Avicennia germinans (Photo credit A.R. Armitage)
Aerial roots that extend from the mangrove trunk are termed prop roots, and
those that protrude upward from lateral belowground roots are called pneumatophores (Fig. 13). The aerial portions of these roots are covered with large
pores called lenticels. Air is taken up through the lenticels and transported through
the aerenchyma tissue to the belowground root system (Fig. 11), thus delivering
the oxygen necessary for root cellular metabolism in otherwise hypoxic or
anoxic soils.
Zonation
In concept, intertidal zonation patterns are dictated by physiological responses of
each species to abiotic stressors that vary along tidal gradients. Mangroves are
somewhat plastic in their internal and external morphology, so some species can
occur at a range of elevations, and zonation patterns are variable within and among
geographic regions of the world. A wide variety of factors, including shoreline
topography, tidal and freshwater influence, salinity, and sediment characteristics,
influence mangrove distribution along elevation gradients. Thom (1984) identified
no fewer than eight distinct geomorphic and biological settings that have unique
mangrove zonation patterns. This section will focus on some of the most common
types of mangrove tidal zones, with specific emphasis on the species common to
Caribbean mangrove swamps.
The land-sea interface, often referred to as fringe mangrove habitat, is characterized by permanently flooded soils, giving the plants constant exposure to salt water.
15
441
Fig. 13 Mangrove aerial root structures. (a) Prop roots on a juvenile red mangrove, Rhizophora
mangle. (b) Pneumatophores extending upward from lateral roots of a juvenile black mangrove,
Avicennia germinans (Photo credit A.R. Armitage)
The soils generally have low oxygen content, though they are not necessarily
anoxic (McKee 1993). Oxygenic phototrophs such as diatoms and other eukaryotic
algae inhibit nitrogen fixation, thereby maintaining low soil nitrogen content in
fringe mangrove soils (Fig. 14; Lee and Joye 2006). This aerobic activity also
facilitates sulfide oxidation, reducing the buildup of toxic sulfides (Fig. 14; Sherman et al. 1998). In the Caribbean, the red mangrove (Rhizophora mangle) dominates this fringe habitat. With its characteristic, prominent prop roots (Fig. 13a),
red mangroves form an iconic image of the Caribbean coastline. Prop roots are
covered with lenticels and contain aerenchyma tissue, enabling red mangroves to
survive in permanently flooded soils. Red mangroves also have heavily suberized
roots that can block up to 99 % of salt uptake from the flooding seawater. The long,
thin propagules characteristic of red mangroves (Fig. 12a) are an additional adaptation to the salt water environment.
The zone above the fringe habitat is difficult to succinctly characterize. In some
areas, this zone is called a transition habitat that contains a mix of species. In other
areas, this drier habitat is called a basin habitat and is dominated by just one or two
species. In general, the flooding duration in mid-elevation habitats is relatively
short, facilitating the diffusion of oxygen from the atmosphere into the soils. As in
the fringe habitat, nitrogen and sulfide accumulation rates are relatively low
(Fig. 14). The shorter flood periods allow mangroves in this zone to have somewhat
reduced aerial root structures. In the Caribbean, black mangrove (Avicennia
germinans) is characteristic of this zone. The pneumatophores of this species can
442
A.R. Armitage
Fig. 14 Excerpt from Figs. 1, 3, and 5 in Sherman et al. (1998). Changes in mangrove and soil
characteristics with increasing distance from the shoreline (Reprinted with permission from
Springer-Verlag)
extend upward out of the ground for several meters away from the primary tree
trunk (Fig. 13b). Like prop roots, pneumatophores have aerenchyma and lenticels to
facilitate gas exchange and root aeration. Black mangroves roots are suberized, but
not as heavily as red mangrove roots. Black mangroves manage excess salt uptake
by secreting salt through numerous small salt excretion glands scattered across leaf
surfaces. The production of small but numerous propagules (Fig. 12b) facilitates
seedling survival in saline soils.
15
443
444
A.R. Armitage
Fig. 15 Excerpt from Fig. 4 in Onuf et al. (1977). Mean numbers ( SE) of leaves, branches, and
flowers added per 1-cm diam. main stem in high- (solid line) and low- (dashed line) nutrient areas.
Differences between sites were significant by t-tests (df 10) for dates where *( p < .05) or **
( p < .01) appear in the upper part of the figure (Reprinted with permission from the Ecological
Society of America)
Fig. 16 Simplified
conceptual diagram depicting
the interaction between topdown and bottom-up forces
influencing mangroves on
islands that are used as
rookeries
15
445
tropical coastlines, smaller salt marsh species are outcompeted by dense, tall
mangrove canopies. In some subtropical areas, there is a transition zone an
ecotone between marsh and mangrove habitats. These ecotones occur in temperate areas of Australia, New Zealand, and the southern continental United States.
Mangrove-marsh ecotones are dynamic habitats mangroves often expand into
salt marshes during periods with warm winters and contract during periods with
hard freezes. This dynamic is primarily driven by temperature, but many other
factors influence mangrove-marsh distribution as well, including rainfall, salinity,
sea level, propagule supply, and interspecific competition. For example, Spartina
alterniflora can outcompete newly sprouted black mangrove propagules (McKee
and Rooth 2008), but if the mangrove seedlings survive through a few growing
seasons, the established tree will begin to displace the surrounding marsh grasses
and forbs.
Current research suggests that mangrove distributions may continue to expand
in response to climate change. For example, models predict that an increase in
winter minimum temperatures of 24 C may lead to black mangroves replacing
salt marsh on nearly all of the Texas and Louisiana coastlines by the year 2100
(Fig. 17; Osland et al. 2013). Other climate-related factors that may increase
mangrove expansion rates include rising sea level due to glacial melting and
thermal expansion. As little as 10 cm of sea level rise over the next 100 years will
likely result in substantial mangrove expansion in all Gulf of Mexico states; sea
level rise may cause mangroves to displace over 10,000 ha of coastal marsh in
both Florida and Louisiana (Doyle et al. 2010). Climate change scenarios that
include increasing atmospheric carbon dioxide concentration and changing herbivore populations will also likely influence mangrove-marsh dynamics, though
these interactions are complex. Elevated CO2 alone may not be sufficient for
mangrove seedlings to outcompete marsh plants, but if there is also low herbivory
pressure and sufficient nitrogen supply, then elevated CO2 may accelerate mangrove growth (McKee and Rooth 2008). The exact role of each of these factors,
and how they interact with each other, is a rapidly growing field of study in
coastal plant ecology. Furthermore, appropriate management of coastal resources
depends on our understanding of the ecological implications of this shift in plant
communities. Will this change in plant species composition alter the ecosystem
services that wetlands provide, such as fishery nurseries, erosion control, or water
quality improvement? Key ecosystem services of coastal plant habitats are
described in the next section.
Fig. 17 Excerpted from Fig. 6 in Osland et al. (2013). Predictions of mangrove forest relative abundance (i.e., percentage of tidal saline wetlands dominated
by mangrove forests) under alternative future (20702100) winter climate projections: left panel, mangrove forest relative abundance with an ensemble B1
scenario climate; right panel, mangrove forest relative abundance with an ensemble A2 scenario climate. Note that these predictions apply just to the tidal
saline wetland habitat within each cell and not the entire cell. Climate scenarios are defined by the Intergovernmental Panel on Climate Change (Reprinted
with permission from Blackwell Publishing Ltd.)
446
A.R. Armitage
15
447
factors, including the provision of food for sustenance, monetary gain, aesthetic
value, and clean air and water. Several ecosystem functions of coastal plant
communities that are particularly valued by humankind are highlighted in the
following section.
Water Quality
Coastal plant communities are widely recognized for their capacity to improve
nearshore water quality. This plant-mediated improvement of water quality is
termed phytoremediation. Coastal wetlands are not stagnant water bodies
many have slow but directional water flow from inland sources to nearshore habitat.
Some wetlands are specifically constructed to manage water flow between terrestrial and marine ecosystems these are called treatment wetlands. The plants in
natural and treatment wetlands provide frictional resistance, slowing down water
flow, thus facilitating the removal of nutrients, bacteria, and other pollutants
through a variety of mechanisms. When wetland plants lower water velocity, this
facilitates the settlement of suspended solids and adhered contaminants. Settlement
is the primary mechanism for removal of organic solids (i.e., sewage waste) from
water moving through coastal wetlands. Many nutrients, especially ammonium,
nitrate, and phosphate, can be removed from the water through direct uptake by
plants and bacteria, which then use these nutrients for metabolic processes. Bacteria
in wetland soils can transform ammonium into nitrate (nitrification) and then into
N2 gas (denitrification). Nitrogen gas can then volatilize (evaporate or diffuse)
from the water into the atmosphere. Some nutrients, particularly inorganic forms of
phosphorus, can become tightly bound to clay particles in a process called adsorption. These phosphorus-clay complexes are largely biologically inert, and as the
clay particles settle to the benthos, the phosphorus is functionally removed from the
water column.
448
A.R. Armitage
Coastal wetlands also play an important role in the global carbon cycle, particularly given their potential for carbon sequestration. Carbon sequestration occurs
when carbon assimilation is greater than carbon loss in an ecosystem. In marine
environments, including coastal wetlands, sequestered carbon is referred to as blue
carbon. Mechanisms of carbon assimilation in wetlands include photosynthesis,
soil microbe assimilation, and the decomposition and burial of plant tissue. Carbon
is lost from wetlands through microbial and plant respiration and through the
decomposition and export of plant tissue into adjacent waterways. Natural and
anthropogenic (human-caused) wetland loss can accelerate carbon loss and reduce
sequestration potential. Changes in wetland vegetation such as the shift from
marsh- to mangrove-dominated systems may also alter the blue carbon storage
potential in wetlands; the nature of these potential changes is a currently growing
field of study.
15
449
Nursery Habitat
A wide range of commercially and recreationally important fish and invertebrate
species rely on coastal wetlands, especially salt marshes, for part or all of their life
cycle. In fact, over 75 % of commercially and recreationally targeted fishery
species spend at least part of their life cycle in estuarine wetlands. For example,
red drum (Sciaenops ocellatus) is a popular sport fish on the Atlantic and Gulf
coasts of the United States. This fish spawns in nearshore habitats. Larvae and
juveniles reside in estuaries, foraging on small shrimp, crabs, and other larval fish
in salt marshes at high tide. Shrimp fisheries are also dependent on salt marshes. In
the Gulf of Mexico, brown (Farfantepenaeus aztecus) and white shrimp
(Litopenaeus setiferus) spawn at sea but inhabit Spartina alterniflora or Juncus
spp. marshes in the postlarval (non-planktonic) stage. These shrimp fisheries are
most productive in areas with extensive estuarine marshes, like the Mississippi
Delta.
Recreation
Wetland plants provide habitat for many species of animals beyond those that
directly contribute to commercial fisheries. Many recreationally fished species
also rely on coastal wetlands. In the Gulf of Mexico, for example, over 80 % of
recreationally targeted species spend at least some of their life in estuarine wetlands. Coastal wetlands also provide critical stopover and wintering habitat for
migratory birds: on a typical winter day in any given coastal wetland in Baja
California, 5,000 or more migratory shorebirds may be spotted. Some coastal
450
A.R. Armitage
wetlands provide essential habitat for endangered species, such as the whooping
crane (Grus americana), which forages exclusively in salt marshes in Texas in the
winter. While enjoying these diverse and abundant wildlife populations, recreational fishers and birders contribute billions of dollars to coastal economies each
year. In 2006, a typical year, birders alone contributed to $82 billion in total
industry output to the United States economy, primarily through purchases of
lodging, transportation, food, and equipment (Carver 2009).
Development
Modern civilization, and accompanying urban and agricultural development, has
dramatically altered coastal ecosystem landscapes. Some wetlands are filled for
urban development; other developments occur in upland habitats directly adjacent
to wetlands. The higher elevations of mangrove swamps are sometimes cleared to
create room for urban growth or resort communities. Other mangrove swamps are
cleared and excavated to create room for mariculture ponds to grow shrimp or other
farmed seafood resources. Coastal marshes have been diked or drained to create
agriculture fields or livestock grazing habitat; other marshes are flooded for rice
farming.
In addition to directly causing habitat loss, development also increases groundwater use, which can accelerate subsidence. Subsidence is the gradual lowering of
the sediment surface through mechanisms such as sediment compaction. Natural
subsidence occurs slowly and is usually mitigated by the accumulation of sediment
that enters estuaries from rivers. However, subsidence rates can be greatly exacerbated by anthropogenic activities, especially the withdrawal of groundwater.
A particularly striking example of anthropogenic subsidence was documented around
Houston, Texas, in the 1970s. A booming oil industry spurred population growth in
the area, driving up the industrial and residential demand for groundwater. Rapid
withdrawal of groundwater accelerated subsidence, and over a period of less than
10 years, many neighborhoods sunk more than half a meter. Some localized spots
sank even more up to 3 m (Fig. 18). This rapid subsidence permanently inundated
tidal marshes, causing over 95 % marsh loss in a very short time period.
Entire neighborhoods had to be abandoned due to chronic flooding problems.
15
451
Fig. 18 Google Earth images of Armand Bayou (near Houston, TX) in 1953 and 2012. In the
1953 image, note the tidal marshes in Horsepen Bayou and at marker #1 and the narrow tidal
channel at marker #2. By 2012, subsidence had flooded most of those features. Ongoing restoration
work in Horsepen Bayou is reestablishing some of the tidal marsh features
452
A.R. Armitage
Freshwater Diversion
Recall the concept of an estuary: a body of water where fresh and salt water mixes.
Many estuaries are parts of heavily developed watersheds, which are the areas
encompassing all the lakes and rivers that eventually drain into a large water body.
Demands for fresh water from urban and agricultural developments ultimately
reduce freshwater input to the estuaries. What happens to an estuary when freshwater inflows decrease? The most acute impact, arguably, is an increase in salinity.
These increases in salinity are likely to be exacerbated by extreme environmental
events phenomena like droughts (Fig. 1). Long-term effects of high salinity could
include plant or animal die-offs or shifts toward more marine species assemblages.
Eutrophication
Plant productivity in most ecosystems is limited by particular nutrients those
nutrients that are in shortest supply relative to others, and will therefore limit
organism growth. In pristine coastal habitats, nitrogen and phosphorus are typically the most limiting. There are many anthropogenic sources of these limiting
nutrients, including fertilizer runoff, sewage, and livestock waste. Moderate input
of anthropogenic nutrients can increase ecosystem productivity, but excessive
nutrient input can cause anthropogenic eutrophication: the rapid buildup of
organic matter. In salt marshes, plants respond to excess nutrients by accelerating
aboveground production: this produces the excess organic matter that is characteristic of eutrophic conditions. However, increased aboveground production is typically matched by a decrease in belowground production (Deegan et al. 2012).
Lower root biomass is linked to decreased sediment stability, which eventually
results in marsh erosion and habitat loss.
Policy
Wetlands are currently the only ecosystem with an international agreement focused
on conservation and sustainable utilization. This agreement, the Ramsar Convention, was formed in 1971 by conservation groups in Europe that recognized the
ecological and economic implications of widespread wetland loss. Currently, at
least 163 nations are members of the convention. Central to the Ramsar Convention
is the wise use concept: wetlands should be conserved and sustainably used for
the benefit of humankind. Although the Ramsar Convention has no regulatory
power, it has helped nations identify conservation priorities and define management
strategies.
In the spirit of the Ramsar Convention, George H.W. Bush adopted a No Net
Loss policy for the United States in 1989. The essence of this policy is that for
every one acre of wetlands that is lost, at least one acre must be created or restored
in its place. This policy applies specifically to jurisdictional wetlands, which are
15
453
generally defined as those wetlands that fall under federal or local protection, based
on the 1977 Section 404 amendment to the Clean Water Act (Kruczynski 1990).
The definition of jurisdictional wetlands has narrowed and widened at times in
response to sometimes contentious disputes among landowners, developers, environmental groups, and federal management agencies. These legal scuffles are
complex and ongoing, but at this time, most coastal wetlands, including salt
marshes and mangroves, are protected by the No Net Loss policy.
Restoration
The No Net Loss policy stipulates that if development impacts jurisdictional
wetlands, then an equivalent area of wetland needs to be restored as compensation
for the impact. The process of wetland restoration is simple in concept, but
challenging in practice. In concept, wetland restoration first involves creating
(by excavating, filling, or leveling) an appropriate elevation for the targeted plant
species. Then, plants are allowed to establish naturally or are transplanted into the
site an undertaking that often involves large groups of volunteers, who then
develop a stewardship of the new habitat (Fig. 19). Once plants are established,
the Field of Dreams hypothesis is usually implicitly or explicitly invoked: If you
build it, they will come (Palmer et al. 1997). In this context, they refers to the
454
A.R. Armitage
References
Baxter I, Hosmani PS, Rus A, Lahner B, Borevitz JO, Muthukumar B, Mickelbart MV,
Schreiber L, Franke RB, Salt DE. Root suberin forms an extracellular barrier that affects
water relations and mineral nutrition in Arabidopsis. Plos Genet. 2009;5:e1000492.
15
455
Bertness MD. Ribbed mussels and Spartina alterniflora production in a New England salt marsh.
Ecology. 1984;65:1794807.
Bertness MD. Fiddler crab regulation of Spartina alterniflora production on a New England salt
marsh. Ecology. 1985;66:104255.
Bertness MD. Zonation of Spartina patens and Spartina alterniflora in a New England salt marsh.
Ecology. 1991;72:13848.
Bertness MD. The ecology of a New England salt marsh. Am Sci. 1992;80:2608.
Boston KG. The development of salt pans on tidal marshes, with particular reference to southeastern Australia. J Biogeogr. 1983;10:110.
Brody SD, Highfield WE, Ryu HC, Spanel-Weber L. Examining the relationship between
wetland alteration and watershed flooding in Texas and Florida. Nat Hazards. 2007;
40:41328.
Carver E. Birding in the United States: a demographic and economic analysis. Addendum to the
2006 National Survey of Fishing, Hunting, and Wildlife-Associated Recreation, report 2006-4,
U.S. Fish & Wildlife Service, Arlington; 2009
Craft C. Freshwater input structures soil properties, vertical accretion, and nutrient accumulation
of Georgia and U.S. tidal marshes. Limnol Oceanogr. 2007;52:122030.
Craft C, Reader J, Sacco JN, Broome SW. Twenty-five years of ecosystem development of
constructed Spartina alterniflora (Loisel) marshes. Ecol Appl. 1999;9:140519.
Craft C, Clough J, Ehman J, Joye S, Park R, Pennings S, Guo HY, Machmuller M. Forecasting the
effects of accelerated sea-level rise on tidal marsh ecosystem services. Front Ecol Environ.
2009;7:738.
Dahl TE. Wetlands losses in the United States 1780s to 1980s. Washington, DC: U.S. Department
of the Interior, Fish and Wildlife Service; 1990.
Deegan LA, Johnson DS, Warren RS, Peterson BJ, Fleeger JW, Fagherazzi S, Wollheim
WM. Coastal eutrophication as a driver of salt marsh loss. Nature. 2012;490:38892.
Doyle TW, Krauss KW, Conner WH, From AS. Predicting the retreat and migration of tidal
forests along the northern Gulf of Mexico under sea-level rise. For Ecol Manage. 2010;
259:7707.
Elmqvist T, Cox PA. The evolution of vivipary in flowering plants. Oikos. 1996;77:39.
Gedan KB, Kirwan ML, Wolanski E, Barbier EB, Silliman BR. The present and future role of
coastal wetland vegetation in protecting shorelines: answering recent challenges to the paradigm. Clim Change. 2011;106:729.
Giri C, Ochieng E, Tieszen LL, Zhu Z, Singh A, Loveland T, Masek J, Duke N. Status and
distribution of mangrove forests of the world using earth observation satellite data. Glob Ecol
Biogeogr. 2011;20:1549.
IPCC. Climate Change 2007: the physical science basis. Contribution of working group I to the
Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge,
UK/New York: Cambridge University Press; 2007.
Jordan TE, Valiela I. A nitrogen budget of the ribbed mussel, Geukensia demissa, and its
significance in nitrogen flow in a New England salt marsh. Limnol Oceanogr. 1982;27:7590.
Kruczynski WL. Mitigation and the Section 404 program: a perspective. In: Kusler JA, Kentula
ME, editors. Wetland creation and restoration: the status of the science. Washington, DC:
Island Press; 1990. p. 54954.
Lee RY, Joye SB. Seasonal patterns of nitrogen fixation and denitrification in oceanic mangrove
habitats. Mar Ecol Prog Ser. 2006;307:12741.
McKee KL. Soil physiochemical patterns and mangrove species distribution reciprocal effects? J
Ecol. 1993;81:47787.
McKee KL, Rooth JE. Where temperate meets tropical: multi-factorial effects of elevated CO2,
nitrogen enrichment, and competition on a mangrove-salt marsh community. Glob Chang Biol.
2008;14:97184.
Onuf CP, Teal JM, Valiela I. Interactions of nutrients, plant growth and herbivory in a mangrove
ecosystem. Ecology. 1977;58:51426.
456
A.R. Armitage
Osland MJ, Enwright N, Day RH, Doyle TW. Winter climate change and coastal wetland
foundation species: salt marshes versus mangrove forests in the southeastern U.S. Glob
Chang Biol. 2013;19:148294.
Palmer MA, Ambrose RF, Poff NL. Ecological theory and community restoration ecology. Restor
Ecol. 1997;5:291300.
Pi N, Tam NFY, Wu Y, Wong MH. Root anatomy and spatial pattern of radial oxygen loss of eight
true mangrove species. Aquat Bot. 2009;90:22230.
Sherman RE, Fahey TJ, Howarth RW. Soil-plant interactions in a neotropical mangrove forest:
iron, phosphorus and sulfur dynamics. Oecologia. 1998;115:55363.
Stiven AE, Kuenzler EJ. The response of two salt marsh molluscs, Littorina irrorata and
Geukensia demissa, to field manipulations of density and Spartina litter. Ecol Monogr.
1979;49:15171.
Thom BG. Coastal landforms and geomorphic processes. In: Snedaker SC, Snedaker JG, editors.
The mangrove exosystem: research methods. Paris: Unesco; 1984. p. 317.
Thursby GB, Abdelrhman MA. Growth of the marsh elder Iva frutescens in relation to duration of
tidal flooding. Estuaries. 2004;27:21724.
Vitousek PM, Aber JD, Howarth RW, Likens GE, Matson PA, Schindler DW, Schlesinger WH,
Tilman DG. Human alteration of the global nitrogen cycle: sources and consequences. Ecol
Appl. 1997;7:73750.
Further Reading
Craft C, Megonigal P, Broome S, Stevenson J, Freese R, Cornell J, Zheng L, Sacco J. The pace of
ecosystem development of constructed Spartina alterniflora marshes. Ecol Appl.
2003;13:141732.
Dugan P. Wetlands in danger: a world conservation atlas. New York: Oxford University Press;
1993.
Engle VD. Estimating the provision of wetland services by Gulf of Mexico coastal wetlands.
Wetlands. 2011;31:17993.
Mendelssohn IA, McKee KL, Patrick Jr WH. Oxygen deficiency in Spartina alterniflora roots:
metabolic adaptation to anoxia. Science. 1981;214:43941.
Perry CL, Mendelssohn IA. Ecosystem effects of expanding populations of Avicennia germinans
in a Louisiana salt marsh. Wetlands. 2009;29:396406.
R
utzler K, Feller IC. Caribbean mangrove swamps. Sci Am. 1996;274:949.
Saintilan N, Rogers K, McKee K. Salt marsh-mangrove interactions in Australasia and the
Americas. In: Perillo GME, Wolanski E, Cahoon DR, Brinson MM, editors. Coastal wetlands:
an integrated ecosystem approach. The Netherlands: Elsevier; 2009. p. 85583.
16
Hugh Kirkman
Contents
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Seagrass Ecosystems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Seagrass Morphology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Economic Goods and Services Provided by Seagrass Ecosystems . . . . . . . . . . . . . . . . . . . . . . . .
Hydrodynamics and Resilience in Seagrass Ecosystems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Seagrass Grazers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Epiphytes and Epiphyte Grazers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Complex Food Webs Associated with Seagrass Ecosystems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Threats to the Future Vitality of Seagrass Ecosystems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Restoration and Recovery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Genetic Diversity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Future Directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
458
464
466
466
467
469
469
470
472
476
477
479
480
Abstract
H. Kirkman (*)
Australian Marine Ecology Pty Ltd, Kensington, Australia
e-mail: hughkirkman@ozemail.com.au
# Springer Science+Business Media New York 2014
R.K. Monson (ed.), Ecology and the Environment, The Plant Sciences 8,
DOI 10.1007/978-1-4614-7501-9_20
457
458
H. Kirkman
Introduction
Restoration and remediation of seagrass ecosystems have not met with great
success. The use of vegetative propagules as a means for reestablishment of
seagrass beds has been plagued with difficulties due to mismatches between
propagule sources and targeted restoration beds. Removing vegetative propagules
from donor beds leads to problems of the donor beds recovering. Growing seagrass
from seed is not always a viable option for restoration because of the vulnerability
of seedlings and poor recruitment into unvegetated areas. Remediation of
destroyed seagrass is not often successful. An understanding of levels of genetic
diversity and spatial genetic structure can contribute to improved restoration outcomes by identifying the most genetically appropriate source material for restoration sites. The discoveries made recently through DNA analysis and
phylogenetic affinities have also helped untangle some of the taxonomic identities
of seagrass and led to better decisions as to the choice of restoration sources and
materials.
The ancestors of the higher plants left the sea some 400 million years ago, but the
seagrasses are the only ones to have returned to a completely submerged marine
existence. This polyphyletic group of flowering plants reinvaded the sea probably
about 100 million years ago in the Cretaceous (Larkum and den Hartog 1989). Our
current knowledge of species affinities and phylogenetic origins is poor for this
group of plants and requires urgent improvement in order to better inform management and researchers (Table 1). A stable taxonomy is a necessary base for all
botanical research. Morphological and anatomical variations within the species are
not systematically documented, and it is recommended that samples of material
used for molecular, physiological, and morphological research are deposited in
recognized herbaria.
There are about 60 species of seagrass in the world in 13 genera (Table 1).
Ruppia and Lepilaena are often grouped among the seagrasses but can grow in
brackish and fresh water. There are so few seagrass species globally and locally,
and a large degree of endemism that the loss of one species may mean thousands of
other organisms are lost. Kuo and den Hartog (2001) describe all seagrass to that
date and offer a key for their identification.
16
459
Table 1 List of seagrass species of the world. The distributions have been taken from Green and
Short (2003). The Seagrasses of the World. There is still taxonomic activity deciding on whether
some species here are real species or strains of others. Distributions too are unclear in some cases
Family
Zosteraceae
Genus
Zostera
Phyllospadix
Heterozostera
Cymodoceaceae
Halodule
Species
marina
caespitose
caulescens
asiatica
noltii
japonica
capensis
capricornii
muelleri
mucronata
novazelandica
scouleri
torreyi
serrulatus
iwatensis
japonicus
tasmanica
polychlamis
nigricaulis
chiliensis
uninervis
beaudetti
wrightii
bermudensis
ciliate
pinifolia
emarginata
Cymodocea
nodosa
rotundata
serrulata
angustata
Syringodium
filiforme
isoetifolium
Thalassodendron ciliatum
pachyrhizum
Amphibolis
antarctica
griffithii
Distribution
Europe, North America
Japan
North Korea and Japan
Korea and Japan
East Atlantic, Baltic, Mediterranean,
Black, Caspian, and Aral Seas
Japan
Southern Africa
Australia
Australia
Australia
New Zealand
Western North America
Western North America
Northwestern North America
Korea, China, and Japan
Korea and Japan
Southern Australia
Southern Australia
Southern Australia
Chile
Tropical and subtropical Australia,
West Africa, SE Asia, India, Pacific
Northeast Madagascar, Caribbean
Global
Bermuda
Tobago Island, Panama
Indo-West Pacific
Brazil
Mediterranean and North Africa
Indo-West Pacific
Indo-West Pacific
Northwestern Australia
Caribbean, Florida
Indo-West Pacific
Indo-West Pacific
South Western Australia
Southern Australia
Southern Australia
Southern Australia
South western Australia
(continued)
460
H. Kirkman
Table 1 (continued)
Family
Posidoniaceae
Genus
Posidonia
Hydrocharitaceae Enhalus
Thalassioideae
Thalassia
Halophiloideae
Halophila
Halophila sect.
Microhalophila
Halophila sect.
Spinulosa
Halophila sect.
Tricostatae
Halophila sect.
Americanae
Ruppiaceae
Ruppia
Zannichelliaceae Lepilaena
Species
oceanica
australis
sinuosa
angustifolia
ostenfeldii
robertsoniae
coriacea
denhartogii
kirkmanii
acoroides
hemprichii
testudinum
ovalis
ovata
minor
australis
hawaiiana
madagascariensis
johnsonii
decipiens
capricorni
beccarii
Distribution
Mediterranean
Southern Australia
Southern Australia
Southern Australia
Southern Australia
Southern Australia
Southern Australia
Southern Australia
Southern Australia
Indo-West Pacific and Australia
Australia
Caribbean and Florida
Global
Trop. Australia, Southeast Asia
Australia, SE Asia, Western Pacific
Southern Australia
Hawaii
Madagascar
Florida
Australia
Queensland and New Caledonia
India and SE Asia
spinulosa
tricostata
engelmannii
baillonii
tuberosa
marina
The taxa regarded as seagrasses belong to four families, viz., the Zosteraceae,
the Cymodoceaceae, the Posidoniaceae, and the Hydrocharitaceae. The first three
families contain only seagrasses, but the Hydrocharitaceae contains only three
genera that are considered seagrasses. The other 14 genera are confined to freshwater habitats. Two other families contain one species each, and these have not
received a lot of research Ruppia tuberosa and Lepilaena marina (Table 1). Nine
of the 13 genera are dioecious.
Sculthorpe (1969) gave a very comprehensive description of the morphology,
physiology, and ecology of submerged aquatic plants in his definitive book.
Seagrass plants have adapted to being supported by water and have nonfunctional
stomates; they assimilate dissolved CO2 by diffusion through the epidermis which
16
461
is the major site for photosynthesis, in contrast to terrestrial plants. Seagrasses vary
in their ability to grow in low-light conditions. Most species of seagrass are adapted
to lower light levels, and they have evolved gas storage organs, both of which can
be considered adaptations that allow them to photosynthetically assimilate CO2 at
low, but sufficient rates. Seagrasses have a thin cuticle over the leaf blade and are
halophytic in their physiological traits. Most can live for short periods in a wide
range of salinities; the salinity of coastal seawater is about 35 parts per 1,000.
Seagrasses also withstand a wide range of temperatures in the coastal waters and are
capable of acclimating to seasonal and spatial variability in this environmental
factor. Zostera marina was found to be growing healthily under ice in an embayment of the Bering Sea. Furthermore, it was living there in anaerobic conditions.
Thus, these plants are quite robust in their adaptive potential! Seagrasses have
become anatomically adapted to limited access to oxygen by developing
aerenchymatic tissues with continuous air-filled lacunae running from leaves to
roots. Oxygen is only lost to the water column during the day, but it is continuously
lost from roots and rhizomes to the sediment. The oxygen produced in photosynthesis is stored in lacunal spaces of the leaves and can be recycled for use in a
limited and localized rate of aerobic respiration. The loss of oxygen to the rhizosphere from root surfaces is vital to protect root tissues by oxidizing reduced toxic
phytotoxins like iron, manganese, and sulfide. The oxygen released to sediments
has important implications for the degradation of organic matter, acting as the
terminal electron acceptor in the oxidative breakdown of organic molecules.
Seagrasses may be monoecious or dioecious. Pollination in the seagrasses takes
place in the water column except in Lepilaena and Enhalus where pollen is released
at the surface. In Enhalus the male flower breaks the surface and releases the
floating pollen to the receptive female flower, and a number of seeds mature in a
fruit that may be 510 cm long. The seeds germinate on release (McConichie and
Knox 1989).
Seagrass ecosystems grow in coastal waters from intertidal to 50 m deep or
more. This is an important statement to make at the beginning of a chapter on
seagrass ecosystems. Seagrasses are limited in their distribution by light, and 50 m
is about the limit that suitable light can penetrate even the clearest coastal waters.
Seagrasses require an underwater photosynthetic irradiance more than 11 % of that
incident on the water surface. Light is reduced by turbidity in the water, and this
turbidity is determined by the content of sediment or organic matter. Light is also
reduced by animals or plants growing on the seagrass plants; these epiphytes, as
they are called, can often shade seagrass plants to below the photosynthetic
compensation point required to sustain plants, leading to death under high nutrient
conditions.
Most temperate seagrasses are seasonal having a strong growth in spring and
early summer then declining in productivity in fall and winter. In a Posidonia
australis bed growing in Port Hacking, New South Wales, Australia, the relative
growth rate measured as mg of carbon per gram of leaf per day closely followed
water temperature (Fig. 1). There is a steep increase in relative growth rate at the
beginning of spring to a maximum at the end of summer. When the mean weight of
462
H. Kirkman
Fig. 1 Average relative growth of Posidonia australis leaves from April 1977 to April 1978, with
surface water temperature over the seagrass bed. Vertical lines are standard errors about the mean
(Kirkman and Reid 1979)
Fig. 2 Dry weight biomass of Posidonia australis estimated for a 15-month period. Vertical lines
represent one standard error about the mean (Kirkman and Reid 1979)
leaves and rhizomes were charted separately for 15 months, there was not such a
seasonal influence as there was in productivity (Fig. 2). These biomasses were the
means from ten quadrats each of 0.0625 m2. These records are important because
they represented measurements that could be used for monitoring seagrass condition. Obviously productivity is a more sensitive measurement to detect changes.
Unfortunately these measurements are more difficult to make in the field than
biomass measures, and we found later that, for large-leaved plants, shoot density
16
463
464
H. Kirkman
recovered in North Carolina but not in Chesapeake Bay. In southern Florida, in the
late 1980s to early 1990s, the pink shrimp, Penaeus duorarum, declined by 50 %
when there was a 20 % loss of Thalassia, the main nursery (from Butler and
Jernakoff (1999), Chap. 2).
Seagrass Ecosystems
Seagrasses have diversified and spread to become dominant organisms throughout
the worlds shallow sediment bottoms around all continents except Antarctica,
primarily in estuaries and more sheltered coastal seas. Two genera (the northeastern
Pacific Phyllospadix and the temperate southern sea nymph, Amphibolis) have
even colonized rocky shores. Colonization by seagrasses profoundly changed the
nature of coastal sediment systems.
Aboveground, the often dense vegetation strongly reduces the physical energy of
waves and currents, creating a zone of kinetic stability within which animal
communities can thrive; in addition, it provides food for herbivores and physical
structure that shelters a much higher abundance and diversity of animals than do the
surrounding bare sediments. The refuge value of seagrasses generally rises with its
species or density complexity. Seagrass leaves provide a substratum for growth of
epiphytic microalgae and sessile invertebrates and macroalgae that fuel complex
food webs. This combined productivity of seagrasses and associated algae ranks
seagrass beds among the most productive ecosystems on earth (Table 2).
Moreover, because much seagrass production ends up in belowground tissues
and ungrazed detritus, seagrass beds are an important global sink for carbon,
accounting for an estimated 15 % of net CO2 uptake by marine organisms on a
global scale, despite contributing only 1 % of marine primary production. Tropical
seagrasses tend to support higher metabolic rates and somewhat lower net community production than temperate ones. The production-to-respiration ratio tended to
increase with gross primary production exceeding 1 on average. It has been
estimated that for a low global seagrass coverage of 300,000 km2 from 20 to
50 Mt of carbon per year and for a high seagrass coverage of 600,000 km2 from
40 to 100 Mt of carbon per year (Duarte et al. 2010) has been taken up.
Seagrass beds provide important nursery areas for juvenile fish including commercially and recreationally used fish and shrimp. For example, in the Gulf of
Carpentaria in Northern Australia, juvenile P. esculentus (tiger prawns) live in
seagrass beds and reach sexual maturity at a carapace length of around 32 mm.
Although seagrass biomass in the Gulf of Carpentaria was not a consistent linear
predictor of juvenile tiger prawn numbers, mean catches of both the 22.9 mm
carapace length postlarvae and juvenile P. esculentus were highest when the
biomass of seagrass exceeded 100 g m 2. However, these high-biomass seagrass
beds contribute only 6 % to the total extent of seagrasses in the shallow waters
(<2.5 m deep) of the Gulf of Carpentaria. Although the numbers of juvenile tiger
prawns were lower in the low-biomass seagrass beds, because of their extent, these
seagrass beds are the main nurseries for sustaining the production of the valuable
16
Table 2 Comparison
between average seagrass
and other marine and
terrestrial ecosystems. Net
primary production (NPP)
(Modified from Mateo
et al. (2007))
Ecosystem
Mangroves
Seagrass
Forests
Macroalgae
Crops
Terrestrial
Phytoplankton coastal
Phytoplankton ocean
465
NPP
(gCm 2/year)
1,000
817
400
375
350
200
167
130
466
H. Kirkman
not have the resources available that provide for a stable ecosystem. Nutrient
cycling and availability may not be as concentrated as they were in an entire
unfragmented bed, producing areas that decline below the size that can withstand
storms and wave surges.
Seagrass Morphology
Seagrasses are rooted plants, and many form dense mats of rhizomes in the
underlying sediments which reduce the mobility of those sediments and thus
stabilize components of local biogeochemical cycles. Roots are not usually supportive organs but have root hairs of variable size and density. The roots of
seagrasses are adventitious and grow from the lower surface of the rhizomes,
generally at the nodes. Seagrass rhizomes are usually herbaceous and
monopodially. Monopodial branching occurs when the terminal bud continues to
grow as a central leader shoot and the lateral rhizomes remain subordinate or
irregularly branched; however, in Amphibolis and Thalassodendron the rhizome
branches sympodially and becomes woody. Sympodial branching occurs when the
terminal bud ceases to grow (usually because a terminal flower has formed) and an
axillary bud or buds. Rhizomes are almost always buried in the sediment, and the
persistent fibrous remains of old leaf sheaths usually cover the rhizomes of Enhalus
and Posidonia and partially cover the rhizomes in some other genera. The coverage
of decomposing leaf sheaths on rhizomes likely provides protection from physical
damage as rhizomes are abrased by sediment movement. The leaf is produced either
from the rhizome nodes, normally from the upper side in Enhalus, Posidonia, and
the Zosteraceae, or from the apex of erect stems in Thalassia and the
Cymodoceaceae. The leaf sheath is clearly differentiated from the leaf blade and
encloses the young, developing leaves in all seagrass genera with ribbon blades.
Thalassia and Amphibolis leaves and sheaths abscise together. Leaf sheaths also
provide unique protective microhabitats for small invertebrates and their larvae.
16
467
practically, they vary so widely and have such broad uncertainties when considered
together that it is better to gain specific value estimates for specific sites.
Even at the site scale, there is still a large number of ecosystem services that have
either no or very unreliable valuation estimates. The most significant problem faced
in valuing ecosystem services, including those of seagrasses, is that very few are
marketed. Some of the products arising from seagrasses, such as raw materials,
food, and fish harvests, are bought and sold in markets; it is easiest to place financial
value on these products.
However, the valuation process, even for these products, is more complicated
than it first appears. For example, one important service of seagrass beds is the
maintenance of fisheries through providing coastal breeding and nursery habitat.
Although many fisheries are exploited for commercial harvests sold in domestic
and international markets, studies have shown that the inability to control fishing
access and the presence of production subsidies and other market distortions can
impact harvests, the price of fish sold, and, ultimately, the estimated value of the
seagrass habitat in supporting commercial fisheries (Barbier et al. 2011). There is a
need for more financial models that include higher-order economic connections and
feedbacks in order to more accurately estimate the values of seagrass ecosystems. It
is likely that human behavior in both financial and regulatory arenas will have to be
added to such models, making it crucial that ecologists work with economists and
social scientists to develop novel modeling frameworks.
468
H. Kirkman
emerge from their sheaths). This means that there is no gap for water to flow
through and hence results in a smoothly decreasing water velocity profile. In the
genus Posidonia there are two distinct groups: the australis group and the
ostenfeldii group. The australis group has stout underground rhizomes that grow
laterally in the sediment. This allows them to spread into unvegetated areas, but
they do not have the strong hold on the sediment that is exhibited in the ostenfeldii
group. This group can grow in strong swells and has a typical windrow appearance
due to the fact that its seedlings only grow successfully on the lee side of sand
ripples. When establishing, the seedlings of members of this group grow as a clump
because their rhizomes grow downward once they have established on the lee side
of the sand ripples, unlike the lateral pattern of growth in the australis group.
Gradually the clump enlarges until it coalesces with others, and a full cover is
achieved. The leaves of this group are also noticeably stronger than those of the
australis group.
Exposure to hydrodynamic energy is widely considered an important environmental factor influencing seagrass species distributions; however, its influence
compared to other mechanisms has not been tested in many places, and this
generalization needs broader consideration. Recently, Hansen and Reidenbach
(2013) have shown the importance of Zostera marina in reducing velocities of
water over them by 60 % in the summer, when leaves were longer, and 40 % in
winter compared with an unvegetated site. The seagrass bed also dampened wave
heights in all seasons except winter when leaves were shortest. Shear stress was
reduced in the summer so that less sediment was resuspended and plants had more
light for photosynthesis. Suspended sediment was enhanced by low seagrass coverage in winter compared with an unvegetated site.
Hydrodynamic processes also influence the dispersal of seagrass seeds and
vegetative fragments, as well as eggs and larvae of organisms that inhabit seagrass
communities and form associated food webs, e.g., invertebrates and fish. Seagrasses
baffle unidirectional tidal and oscillatory (wave-driven) currents. Plant morphology
and structure affect the capacity of seagrasses to influence water flow. The capacity
of seagrasses to baffle water flow and currents is linked to the accretion of
sediments and increases with increasing patch structure and size. This, in turn,
improves conditions for seagrass growth and recruitment, accelerating patch density and the extent of coverage. Empirical studies of temperate seagrass responses
to hydrodynamics, however, have been limited to Posidonia spp. and Amphibolis
spp. in Australia and Zostera marina in temperate USA and Europe. There is room
for much broader consideration of these potential adaptations and influences on
multi-trophic dynamics.
Tidal height and range influence variability in biomass and productivity in
intertidal seagrass populations, e.g., those of Zostera muelleri in Victoria, Southern
Australia, and Halophila decipiens and Halodule uninervis in turbid tropical waters
of great tidal range. Low water levels (tidal heights), barometric conditions, and
high temperatures can prompt prolonged atmospheric exposure and desiccation for
intertidal species which may result in dieback (Seddon et al. 2000). Empirical
studies on the response of seagrasses to atmospheric exposure are limited.
16
469
Seagrass Grazers
Waterfowl are significant grazers of seagrasses consuming large amounts of rhizomes and leaves. Swans (Cygnus atratus) in Australia eat Zostera muelleri while
migratory herbivores such as brant geese (Branta bernicla) live between the
Atlantic coast of the USA from Maine to Georgia, in Alaska, California, and
Mexico and feed on seagrass. In the Gulf of Mexico redhead duck (Aythya americana) eats Halodule wrightii. Swans ingest the rhizomes and leave the leaves to
float off, thus affecting spatial patterns of decomposition. Dugongs (Dugong dugon)
pull out the small plants of Halophila, Cymodocea, and Halodule and, in Shark Bay,
Western Australia, eat Amphibolis antarctica. Dugongs leave circuitous trails in
seagrass beds they have grazed, once again producing the potential for unique
spatial patterning in community and ecosystem processes; this is considered as
the possible basis for ecological interactions and stimulates seagrass growth. The
green sea turtle, Chelonia mydas, eats seagrass and macroalgae in tropical seas.
They tend to graze in grazing plots of Thalassia testudinum in the Bahamas
choosing young leaves by consistent cropping. There is more digestible forage
higher in protein and lower in lignins than ungrazed older leaves. Small fish may
eat seagrass leaves, fruit, and seeds, and some small grazers, such as snails, and
amphipods eat leaf tissue. Because the assimilation rate is quite low, large amounts
are returned as detritus and broken down by bacteria. This interaction of vertebrates, invertebrates, bacteria, and seagrass will affect seagrass growth patterns.
Some invertebrates ingest seagrass leaves, for example, leaf mining linseed isopods
were found in Posidonia leaves with more than 90 % of leaves containing burrows.
The isopods consumed mesophyll tissue and cells of the vascular bundles (Brearley
and Walker 1995).
470
H. Kirkman
Fig. 3 Posidonia australis fruits, note the epiphytes of macroalgae and calcareous polychaetes on
the healthy seagrass leaves (Photograph: H. Kirkman)
16
471
invertebrates that are camouflaged, e.g., isopods, snails, and nudibranchs. Juvenile
shrimp use seagrass beds as nursery areas, and the tiger prawn (Penaeus monodon)
in the Gulf of Carpentaria in Australia is only caught in seagrass beds.
The effects of overfishing on seagrass beds can be quite devastating. A top-down
trophic cascade can occur when the top-level predators are removed. The decline in
large predators brought about by fishing causes an increase in small-fish predators
which deplete populations of mollusc and crustacean grazers that normally reduce
epiphyte loads. Thus, excessive fishing of some upper trophic level fish has the
potential to cause cascading effects down the food web, which ultimately decrease
productivity in the primary producers. This process may have more steps in a
complex food web, but the end result is that seagrass leaves are smothered by
epiphytes reducing the light falling on the seagrass leaves, and if the available light
falls below the compensation point (the light level required to sustain a positive
carbon balance in the plant), the plants will eventually die (Heck and Valentine
2007). The threat of a trophic cascade caused by recreational and commercial
fishing should always be considered.
Under pristine conditions, the older the leaves the more epiphytes there are. In
temperate regions, plants like Posidonia and Amphibolis, which have longer leaf
retention times, may hold more epiphytes than the shorter-lived leaves of
Halophila. Similarly, in the tropics Enhalus will hold more epiphytes than
Halodule or Halophila.
The prolific diversity and abundance of motile, epibenthic, invertebrate fauna
found in seagrass beds can be illustrated by beam trawls through the seagrass at
night when the animals are above the substrate (Fig. 5). A beam trawl for this
purpose is usually a meter wide with a roller to prevent damage to seagrass and has
skids to move it easily over the seagrass vegetation. The net is usually 2 mm with a
472
H. Kirkman
Fig. 5 The animals from a 50 m beam trawl through a Posidonia australis bed at Kangaroo Island
in South Australia (Photograph: H. Kirkman)
1 mm cod end. The beam trawl is pulled along the bottom at about 23 km/h for
50 m collecting all the animals from 50 m2. An example of the difference between
the abundance and diversity of epibenthic fauna in seagrass and on unvegetated
sediment was shown in the Albany harbors in Western Australia. In Princess Royal
Harbour 18, 50 m beam trawl samples on unvegetated sand caught 258 individuals
from 23 species, whereas nearby, in a Posidonia australis bed, 3,923 individuals
were caught from 68 species (Kirkman et al. 1991). The species collected were
amphipods, fish, isopods, molluscs including octopus and squid, and sea cucumbers, brittle stars, and starfish in the echinoderms.
The effect of human impacts on food webs is described by Coll et al. (2011) for
temperate Atlantic seagrass beds. They found that the food-web structure was
similar among low-impact sites in Eastern Canada and a tropical seagrass web
suggesting consistent food-web characteristics across seagrass ecosystems at different latitudes.
16
473
be used to track ecosystem vitality over time. Either the epiphytes can be monitored
regularly or the border of the seagrass bed can be progressively marked and
recorded. These simple measurements at the outer and inner boundaries of the
seagrass bed will assist with management.
Human impacts on seagrasses are well discussed in Ralph et al. (2007). Runoff
from land clearing in preparation for housing and urban construction may be the
largest impact on offshore seagrass meadows. The problem is that the land is
cleared for building and sometimes heavy rains wash off the topsoil because it is
no longer held by vegetation. New roads and cuttings for roads are another source of
sediment runoff. Both of these influences will affect water turbidity and the
potential for seagrass growth by threatening light penetration to the seagrass beds.
In Western Port, Victoria, Australia, beds of the subtidal Zostera muelleri have
been progressively reduced in coverage for the past 50 years. The causes are
difficult to remediate. Erosion from clay cliffs and the shore generally and runoff
from streams and drains have put sediment into the water column. The continual
loss of seagrass has given rise to larger areas of unvegetated mud which is disturbed
in rough weather thus adding to the suspended solids and increasing turbidity.
Reducing erosion from the cliffs is expensive. Terrestrial runoff is due to poor
farming practices and considerable urban development in the catchment and the
loss of vegetated stabilized area continues to exacerbate the problem. Attempts are
being made to grow mangrove as a sediment stabilizer outside the boundary of
seagrass beds and thus reduce wave energy causing erosion.
Development of the coast by building causeways and shoreline armoring may
divert water and generally destabilize beaches and shorelines. Rivers are often
diverted or changed to enable the extraction of freshwater, and this may have an
effect on seagrass beds by favoring one species that prefers seawater (Heterozostera
tasmanica) over Zostera muelleri that has adapted to changed salinity conditions.
Physical damage to seagrass beds can occur when marinas, jetties, and boat
ramps are built on or adjacent to seagrass beds, or these structures may change the
dynamic hydrology (water circulation patterns) of the area, reducing onshore drift
and water flow. Onshore drift is the gradual lateral shift along a beach of beach
material resulting from waves meeting the shore at an oblique angle. Mining or oil
and gas extraction from under seagrass beds are potentially damaging when considering freshwater flows, oil spills, and mining accidents that cause collapse of
mined areas. Moorings and boat ramps add further problems for seagrass ecosystems. The moorings cut spheres in the seagrass bed by chain movement caused by
tides and wind. Boat ramps lead to channels being cut in seagrass beds by boat
propellers at low tides when boats are leaving or returning to the ramp. Adequate
channel markers and a channel will help to prevent this. The main problem with
propeller scouring is that during tidal cycles water washes in and out through these
rills and these are eroded to form quite large channels in which seagrass propagules
are prevented from colonization.
Human occupation of the coastal zone is accompanied by increased rates of
pollution. Industrial chemicals from factories, including heavy metals, petrochemicals, and toxic compounds, are a danger to seagrass ecosystems. These pollutants
474
H. Kirkman
enter the sea from runoff and storm water drains. Agricultural runoff containing
herbicides and insecticides can damage seagrass beds and its associated fauna.
By far the most damaging pollutant in seagrass beds is nutrients. These nutrients
promote epiphyte growth that smothers the photosynthetic potential of seagrasses
and reduces dissolved oxygen levels to dangerously low levels. In marine systems
nitrogen excess is usually the primary culprit. Eutrophication occurs when high
nutrient loads, particularly inorganic nitrogen, are taken up by opportunistic
macroalgae growing on seagrass leaves. The epiphytes and dead seagrass leaves
fall to the substrate and are broken down by bacteria that use up oxygen, and this
anoxic sediment gives off hydrogen sulfide that kills the benthic flora. The whole
seagrass ecosystem may then collapse. Food-web structure and functioning of
seagrass habitats change with human impacts, and the spatial scale of food-web
analysis is critical for determining results (Coll et al. 2011). The spatial scale is a
relevant issue in food-web ecology in general as food webs are typically assembled
in aggregated forms (cumulative or summary webs) due to limited data availability
on trophic interactions.
Dredging near seagrass beds increases turbidity, and this may cause a smothering effect as well, especially if silt screens are not used. If the sediment load is very
high, the effect of seagrass leaves slowing the surrounding water will cause the
sediment to drop out of the water column and smother plants. Dredging should
generally be carried out in the season when seagrass is least productive, for
example, in temperate regions in winter, after carbohydrates and stored material
have been laid down in rhizomes or, in the tropics, in the wet season when seagrass
beds may die out due to low light because of high sediment loads caused by
terrestrial runoff and disturbance of the substrate. They recover naturally during
the dry season.
Globally, disease in seagrasses has not been identified as a major threat. After the
dramatic reduction of the seagrass Zostera marina in the 1930s in the USA and
Europe, recovery was slow and only occasionally has Labyrinthula zosterae, a
marine slime mold-like protist been shown to cause large-scale losses. The death of
seagrasses was attributed to Labyrinthula zosterae, but later it was established that
the plants were under stress and the disease proliferated because of the low
resistance of the seagrasses. Diligent monitoring of seagrass beds will alert managers to conditions that could foster secondary impacts due to disease.
Many of the seagrass beds in the USA and Europe provided insulation material
from the leaves of Zostera marina in the 1920s. The dried leaves, usually recovered
from drift on beaches, were used as insulation in sleeping bags and the walls of
houses. Collections of large amounts of drift material may affect the nutrient
recycling of seagrass beds. There are numerous reports of the slow rate at which
seagrass beds will recover from disturbance. One of these is in Spencer Gulf in
South Australia where Posidonia australis plants were removed to obtain the
underlying fiber. This fiber was from the persistent fibrous remains of old leaf
sheaths of P. australis and was used in clothing manufacture and for insulation in
refrigeration units and steam-heating systems. It is of interest to note that although
this mining was discontinued in the 1920s, the scars where dredges removed the
16
475
fiber are still visible today. This and other evidence from seismic blasting suggests
that Posidonia spp. beds take decades to recover.
Invasive species are a problem in seagrass meadows in some parts of the world.
Of particular note is the damage done by Caulerpa taxifolia in Posidonia oceanica
seagrass beds in the Mediterranean Sea (Meinesz, et al. 1993). Some consideration
should be given to other invasive species that may arrive, e.g., Undaria and Asterias
are potential invaders that could pose problems in the future. Undaria pinnatifida is
an edible kelp called wakame, from Japan, that has invaded seagrass beds and rocky
temperate reefs. Asterias amurensis is the Northern Pacific seastar also from Japan
that removes all organisms from reefs and is also found in seagrass beds.
The full extent of climate change effects on seagrass ecosystems has not yet been
demonstrated or predicted. However, given the changes that have been noted to date
in ocean temperature, salinity, acidification and aragonite saturation, sea level,
circulation, productivity, and exposure to damaging UV light, we can anticipate
significant degradative effects due to climate change in the future. Loss of seagrass
coverage due to exposure to extremes in sunlight or heat has recently been shown in
South Australia (Seddon et al. 2000).
Indirect effects of climate change on seagrass communities could occur due to
intensification and increases in the frequency of tropical and subtropical cyclones.
As discussed above, storms stir up sediment in shallow seas and hence reduce light
to seagrass. Increased storm frequency means that there will be increased turbidity
and this may reduce light to lower than compensation levels for marginal meadows
at the deeper edge. Increased frequency of storms may also disturb seed beds that
normally lie in the sediment, e.g., Halophila ovalis and Halodule uninervis were
lost from Hervey Bay, Queensland, when two very large storms followed each
other, the first destroying the seagrass and the second destroying newly germinated
seedlings (Preen et al. 1995). It took about 5 years for the area to recover. More
intense storms will also increase erosion of edges.
Warmer temperatures and ice cap melting are expected to raise sea levels. For
seagrasses this will bring their habitats shoreward. Those seagrasses growing at the
deeper edge of their habitat may be lost while the shallower margins will gain
coverage. The problem is if development has used those shallower edges to the
point that the seagrass can move no further up the shore, large areas will be lost. The
building of sea walls, coastal roads, housing to the edge of the sea, and other
development must be carefully managed with sea-level rise in mind.
Little is known about the effect of seawater temperature rising, but shifts in
distribution are expected. Seagrass plants cannot move as can some invertebrates
and fish as the water temperature increases. The success of a slow distributional
shift will depend upon the suitability of a new habitat being available, the connectivity between seagrass beds and potential new growth areas, and the dispersal
mechanisms of the propagules.
As carbon dioxide rises in the atmosphere, more is dissolved in seawater leading
to ocean acidification. In seagrass ecosystems, calcareous epiphytes will be the
main victims. The response of calcareous epiphytes to a fall in pH from 8.2
(seawater) to 7.7 in aquaria was a loss of all calcareous algae, and the only calcifers
476
H. Kirkman
were bryozoans at pH 7.7 (Martin et al. 2008). This result may have dramatic
effects on biogeochemical cycling of carbon and carbonate in coastal ecosystems
dominated by seagrass beds.
16
477
Genetic Diversity
The poor knowledge of the minimal habitat requirements for seagrass growth,
colonization and establishment mechanisms, genetic diversity, and reproductive
478
H. Kirkman
modes required to maintain ecologically successful populations hinders the development of sound management practices. The development of molecular DNA
sequencing techniques over the last decade has provided new tools to examine
genetic variability within and among seagrass populations. Much of the power
inherent in molecular genetic data can be tapped, revealing otherwise unobtainable
information at all levels of biotic hierarchy (Kendrick et al. 2005).
Alberte et al. (1994) assisted with breakthroughs in determining that populations
that were morphologically distinct and may have shown different depth distributions could be distinguished by DNA fingerprinting. They also determined that
Zostera marina, in particular, was not characterized by a high degree of clonal
reproduction at spatial scales over 5 m, and they found that Z. marina growing in a
physically disturbed bay had reduced genetic diversity. Knowing the effect that
disturbance has on genetic stability can help establish mitigation and restoration
criteria.
Genetic diversity in terms of greater numbers of distinct clones was positively
associated with seagrass bed density, and this in turn was correlated with greater
invertebrate density, nitrogen retention, and areal productivity. Higher abundances
of invertebrates associated with seagrasses in more genetically diverse Zostera
plots and the positive effects of seagrass genotypic diversity on both seagrass and
grazer biomass depended on grazer species identity. Since mesograzers can have
strong effects on the biomass of both epiphytic algae and seagrasses, and since
seagrass genotypes vary in palatability, understanding the implications of changing
diversity in seagrass ecosystems will require more detailed study of genetic and
species diversity effects at multiple trophic levels. Nevertheless, the picture emerging from controlled experiments and seagrass restoration projects appears consistent: seagrass genetic diversity may be a key variable influencing seagrass
productivity and community processes (Duffy et al. 2013).
There is also a positive impact of clonal diversity along an entire depth gradient
on food-web complexity and density and nutrient retention. Ecosystem restoration
will significantly benefit from obtaining sources (transplants and seeds) of high
genetic diversity and from restoration techniques that can maintain that high genetic
diversity (Reynolds et al. 2012).
Seagrasses provide convincing examples of the broader ecological importance
of genetic or genotypic diversity. Higher allelic diversity within individuals
increased vegetative shoot production and sexual reproduction in transplanted
seagrasses, and transplant success correlated positively with the genetic diversity
of individuals in the source population (Procaccini et al. 2007). More convincing
was the evidence from experimental manipulations of the number of seagrass
genotypes (as measured by DNA microsatellites), which demonstrated that genetic
diversity within a patch can influence primary and secondary production, particularly in the face of disturbance or stress. Patches of eelgrass (Zostera marina) with
greater numbers of clonal genotypes were more resistant to seasonal grazing by
migratory geese, resulting in increased shoot density after grazing in high-diversity
areas and quicker recovery to pre-grazing densities, in the more diverse areas.
Genotypic (and thus phenotypic) diversity also increased the rate of recovery
16
479
from extremely high water temperatures in Zostera marina suggesting that this
effect may be a generalized response to aboveground biomass removal. Subsequent
manipulations that controlled for disturbance confirmed the positive effects of
genetic diversity in the presence and absence of disturbance. Thus there is growing
evidence, albeit only from Zostera so far, that genetic diversity within seagrass
species can be important in buffering seagrasses from several types of perturbations. Genotypic diversity can have positive consequences at the community level
as well.
It is only recently that one has begun to understand the genetics of seagrass
plants and what a seagrass plant is. In Western Australia vast beds of Posidonia
extend for kilometers along the coast; until now it has not been possible to say how
extensive a single plant is. Posidonia oceanica in the Mediterranean is one of the
largest, slowest growing, and longest-lived plants terrestrially or in the sea. In a
recent genetic study of 40 P. oceanica populations across the Mediterranean,
Arnaud-Haond et al. (2012) found individual clones spanning up to 15 km. Based
on the plants known growth rate, such individuals are likely to be thousands,
possibly tens of thousands of years old. This was different from the high degree
of clonal reproduction in Zostera marina shown by Alberte et al. (1994).
The discoveries made by DNA have also helped untangle some of the taxonomic
identities of seagrass. It is at this point that an understanding of levels of genetic
diversity and spatial genetic structure can contribute to improved restoration outcomes. Identifying the most genetically appropriate source material for restoration
sites can be carried out with DNA analysis.
From molecular studies in combination with ecological and hydrological assessments, it is evident that seagrasses are resilient and have persisted in a physiologically challenging submerged environment because they have broad niches. That
local persistence of seagrasses has been achieved by clonal growth and by recruitment from sexually derived propagules. Some seagrasses invest significant amounts
of energy in sexual reproduction, producing seeds with a high capacity for longdistance dispersal that enables them to colonize distant new locations (Kendrick
et al. 2012).
Future Directions
There is a recent trend for widespread loss in tropical and temperate seagrass
ecosystems. Large-scale declines have been reported by Hemminga and Duarte
(2000) at 40 locations, 70 % of which are attributed to human induced disturbance.
There are some areas that have recovered but the long-term trend is for continual
global loss. Short and Wyllie-Echeverria (1996) estimated the area of seagrass lost
globally at 12,000 km2 or about 2 % of the area originally covered. Present losses
are expected to accelerate, particularly in areas of Southeast Asia and the Caribbean
where human pressure is greatest and development incentive is greater than environmental conservation. Restoration of seagrasses seems to be the greatest challenge facing ecologists. Efforts to restore seagrass need to be based on knowledge
480
H. Kirkman
of local conditions, the ecological state of the system prior to disturbance, and
informed decisions about what should be there after restoration. The genetic
investigations into clonal seagrass identity may be helpful in restoration efforts.
It is difficult to separate natural variability from human-caused disturbance. The
role of disturbance and the response by seagrass species to a particular disturbance
should be a major focus of long- and short-term research. Now that climate change
is a component of disturbance, the investigation has become even more complex. It
is recommended that monitoring of seagrass to distinguish between these causes
and to answer relevant questions on management of seagrass ecosystems be
carried out.
As concern increases for the state of natural resources and the degradation of the
worlds oceans, it is critical for countries to progress with conservation actions
specifically focused on seagrass ecosystems. Guidelines for Applying the IUCN
Protected Area Management Categories to Marine Protected Areas (MPA) aim to
make clear what is most significant and of highest priority, and this effort will help
countries more accurately detail their successes (www.iucn.org/pa_guidelines).
These guidelines will define MPAs thus preventing the trend of fisheries advisory
bodies claiming that area mechanisms exploiting fish are MPAs. About 50 % of
global MPAs are considered to have been wrongly allocated because the name of the
MPA, e.g., National Park and Sanctuary, has been used to determine the category,
rather than the management objectives. Confusion tends to arise when sites have
been incorrectly assigned on the basis of activities that occur, rather than using the
stated management objectives. In recent years pressure to deliver success stories has
resulted in false claims of large areas of seagrass being properly protected. It is time
to be realistic about our definition of MPAs in seagrass ecosystems.
Protecting seagrass beds through education of local communities and fishers and
by regulations and even enforcement will help conserve this valuable resource.
Properly regulated marine protected areas will assist with conserving seagrass
ecosystems with benefits to conserving biological diversity and spillover advantages to nonprotected areas.
It is time to stop pretending more areas of seagrass are protected than they
actually are. Understanding which seagrass beds are protected and how they are
protected is of paramount importance in promoting driving global conservation
efforts. Without this information it is difficult to hold the process of determining
marine protected areas in seagrass ecosystems accountable.
References
Alberte RS, Suba GK, Procaccini G, Zimmerman RC, Fain SR. Assessment of genetic diversity of
seagrass populations using DNA FINGER printing: implications for population stability and
management. Proc Natl Acad Sci USA. 1994;91:104953.
Arnaud-Haond S, Duarte CM, Diaz-Almela E, Marba N, Sintes T, Serrae EA. Implications of
extreme life span in clonal organisms: millenary clones in the threatened seagrass Posidonia
oceanic. PLoS ONE. 2012;7(2):e30454.
16
481
Barbier EB, Hacker SD, Chris Kennedy C, Koch EW, Stier AC, Silliman BR. The value of
estuarine and coastal ecosystem services. Ecol Monogr. 2011;81(2):16993.
Borowitzka MA, Lethbridge RC, Charlton L. Species richness, spatial distribution and colonization pattern of algal and invertebrate epiphytes on the seagrass Amphibolis griffithii. Mar Ecol
Prog Ser. 1990;64:28191.
Brearley A, Walker DI. Isopod miners in the leaves of two Western Australian Posidonia species.
Aquat Bot. 1995;52:16381.
Carruthers TJB, Dennison WC, Longstaff BJ, Waycott M, Abal EG, McKenzie LJ, Lee Long
WJ. Seagrass habitats of north east Australia: models of key processes and controls. Bull Mar
Sci. 2002;73(3):115369.
Coll M, Schmidt A, Romanuk T, Lotze HK. Food-web structure of seagrass communities across
different spatial scales and human impacts. PLoS ONE. 2011;6(7):113.
Duarte CM, Marba N, Gacia E, Fourqurean JW, Beggins J, Barron C, Apostolaki ET. Seagrass
community metabolism: assessing the carbon sink capacity of seagrass meadows. Glob
Biochem Cycles. 2010;24:GB4032.
Duffy JE, Hughes AR, Moksnes P-O. Ecology of seagrass communities. Sunderland: Sinaur
Associates; 2013. p. 27197.
Green EP Short FT. World Atlas of Seagrasses Prepared by the UNEP World Conservation
Monitoring Centre, University of California Press, Berkeley, USA. 2003. pp 298.
Hansen JCR, Reidenbach MA. Seasonal growth and senescence of a Zostera marina seagrass
meadow alters wave-dominated flow and sediment suspension within a coastal bay. Estuar
Coasts. 2013;36:1099114.
Heck Jr KL, Valentine JF. The primacy of top-down effects in shallow benthic ecosystems. Estuar
Coasts. 2007;30(3):37181.
Hemminga M, Duarte CM. Seagrass ecology. Cambridge, UK: Cambridge University Press; 2000.
298 pp.
Jernakoff P, Brearley A, Nielsen J. Factors affecting grazer-epiphyte interactions in temperate
seagrass meadows. Oceanogr Mar Biol Annu Rev. 1996;34:10962.
Kendrick GA, Marba N, Duarte CM. Modelling formation of complex topography by the seagrass
Posidonia oceanic. Estuar Coast Shelf Sci. 2005;65:71725.
Kendrick GA, Waycott M, Carruthers TGB, Cambridge ML, Hovey R, Krauss SL, Lavery PS, Les
DH, Lowe RJ, Mascaro O, Vidal OM, Ooi JLS, Orth RJ, Rivers DO, Ruiz-Montoya L,
Statton J, van Dijk JK. and J. Verduin, J.J. The central role of dispersal in the maintenance
and persistence of seagrass populations. BioScience. 2012;62(1):5665.
Kirkman H, Reid DD. A study of the role of a seagrass Posidonia australis in the carbon budget of
an estuary. Aquat Bot. 1979;7:17383.
Kirkman H, Humphries P, Manning R. The epibenthic fauna of seagrass beds and bare sand in
Princess Royal Harbour and King George Sound, south-western Australia. In: Wells FE,
Walker DI, Kirkman H, Lethbridge R, editors. Proceedings of the Third International Marine
Ecological Workshop: The Marine Flora and Fauna of Albany, Western Australia; Perth:
Western Australian Museum; 1991. p. 55363.
Kuo J, den Hartog C. Seagrass taxonomy and identification key. In: Short FT, Coles RG, editors.
Global seagrass research methods. Amsterdam: Elsevier; 2001. p. 3158.
Larkum AWD, Den Hartog C. Evolution and biogeography of seagrasses. In: Larkum AWD,
McComb AJ, Shepherd SA, editors. Biology of seagrasses a treatise on the biology of
seagrasses with special reference to the Australian region. Amsterdam: Elsevier; 1989.
p. 11356.
Martin S, Rodolfo-Metalpa R, Ransome E, Rowley S, Buia M-C, Gattuso J-P, Hall-Spencer
J. Effects of naturally acidified seawater on seagrass calcareous epibionts. Biol Lett R Soc.
2008;4(6):68992.
Mateo MA, Cebrin J, Dunton K, Mutchler T. In: Larkum AWD, Orth RJ, Duarte CM, editors.
Seagrasses: biology, ecology and conservation. Carbon Flux in Seagrass Ecosystems. Dordrecht: Springer; 2007. p. 15992.
482
H. Kirkman
McConichie CA, Knox RB. In: Larkum AWD, McComb AJ, Shepherd SA, editors. Biology of
seagrasses a treatise on the biology of seagrasses with special reference to the Australian
region. Pollination and Reproductive Biology og Seagrasses. Amsterdam: Elsevier; 1989.
p. 74111.
Meinesz A, de Vaugelas J, Hesse B, Mari X. Spread of the introduced tropical marine alga
Caulerpa taxifolia in northern Mediterranean waters. J Appl Phycol. 1993;5:1417.
Preen AR, Lee Long WJ, Coles RG. Flood and cyclone related loss, and partial recovery, of more
than 100 km2 of seagrass in Hervey Bay, Queensland, Australia. Aquat Bot. 1995;52:317.
Procaccini G, Olsen JL, Reusch TBH. Contribution of genetics and genomics to seagrass biology
and conservation. J Exp Mar Biol Ecol. 2007;350:23459.
Ralph PJ, Durako MJ, Enrquez S, Collier CJ, Doblin MA. Impact of light limitation on seagrasses.
J Exp Mar Biol Ecol. 2007;350:17693.
Reynolds LK, McGlathery KJ, Waycott M. Genetic diversity enhances restoration success by
augmenting ecosystem services. PLoS ONE. 2012;7(6):17.
Sculthorpe CD. The biology of aquatic vascular plants. London: Edward Arnold; 1969.
Seddon S, Connolly RM, Edyvane KS. Large-scale seagrass dieback in northern Spencer Gulf,
South Australia. Aquat Bot. 2000;66:297310.
Short FT, Wyllie-Escheverria S. Natural and human induced disturbance of seagrasses. Environ
Conserv. 1996;23:1727.
Further Reading
Butler A, Jernakoff P. Seagrass in Australia: strategic review and development of an R. and
D. plan. Collingwood: CSIRO Publishing; 1999. 210 pp.
Connell SD, Gillanders BM, editors. Marine ecology. Melbourne: Oxford University Press; 2007.
p. 595630.
Duarte CM, Chiscano CL. Above ground and below ground seagrass biomass vs degrees of
latitude. Aquat Bot. 1999;65:15974.
Duffy JE. Biodiversity and the functioning of seagrass ecosystems. Mar Ecol Prog Ser.
2006;311:23350.
Larkum AWD, Orth RJ, Duarte CM. Seagrasses: biology, ecology and conservation. Dordrecht:
Springer; 2007. 691 pp.
Short FT, Coles RG. Global seagrass research methods. Amsterdam: Elsevier; 2001. 473 pp.
Waycott M, Duarte CM, Carruthers TJB, Orth RJ, Dennison WC, Olyarnike S, Calladinea A,
Fourqurean JW, Heck Jr KL, Hughes AR, Kendrick ARG, Kenworthy WJ, Short FT, Williams
SL. Accelerating loss of seagrasses across the globe threatens coastal ecosystems. Proc Natl
Acad Sci USA. 2009;106(30):1237781.
17
Contents
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Phytoplankton Diversity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Picophytoplankton . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Nanophytoplankton . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Microphytoplankton . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Plankton Functional Groups and Trait-Based Phytoplankton Ecology . . . . . . . . . . . . . . . . . . . . . . .
Ecological Roles and Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Emergent Biogeochemical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Trait-Based Phytoplankton Ecology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Characteristics of the Pelagic Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Temperature, Salinity, and Density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Vertical Light Attenuation and Ocean Color . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Vertical Stratification and Mixing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Vertical Nutrient Distributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Dissolved Inorganic Carbon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Ocean Acidification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Primary Production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Net and Gross Primary Production of Marine Phytoplankton . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Net Community and Export Production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Is the Oligotrophic Ocean Autotrophic? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Remote Sensing of Primary Production . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
485
486
489
491
491
493
493
494
495
496
496
499
500
501
502
503
503
505
508
509
509
483
484
511
513
514
516
518
519
521
523
527
529
Abstract
Marine phytoplankton account for about 45 % of global net primary production (NPP). In addition, they perform other important biogeochemical functions including nitrogen fixation, calcium carbonate precipitation, and the
production of climatically active gases such as dimethyl sulfide.
Oceanographers employ a wide variety of platforms for studying marine
phytoplankton ecology, including sampling from ships, sampling from autonomous remotely operated vehicles, and collecting observations from Earthorbiting satellites.
Marine phytoplankton range in size from <1 m in diameter to about 1 mm in
length and include representatives from at least five eukaryotic phyla together
with the cyanobacteria. This wide size range and phylogenetic diversity
presents challenges for quantifying and characterizing phytoplankton
communities.
Functional traits that quantify responses of growth rate, photosynthesis and
nutrient uptake to temperature, irradiance, and nutrient availability provide a
useful basis for understanding phytoplankton ecology.
A variety of complementary approaches are used to measure gross and net
primary production. These include measuring production of O2 and organic
matter in bottle experiments and measuring diel and seasonal changes of O2
in open waters. Information obtained from satellite remote sensing of ocean
color is used to calculate NPP on regional and global scales.
The physical and chemical variables that drive NPP include temperature,
nutrient availability, and solar radiation. These vary in time and space, and
our understanding of this variability is largely encapsulated in the concepts of
the seasonal production cycle and marine biogeochemical provinces.
Nutrient limitation sets an upper limit to NPP over most of the ocean surface,
with either inorganic iron or nitrogen being the proximate limiting element in
different regions.
The upper water column is stably stratified over much of the ocean, and
pronounced vertical gradients of light and nutrients lead to depth separation
of ecotypes with differing adaptations to nutrient availability and the light
environment.
17
485
Introduction
Phytoplankton (from Greek phytos meaning plant and planktos meaning
drifting) are the primary producers that form the base of ocean food webs and
as such play vital roles in ocean ecology. Unlike terrestrial plants, most of which are
macroscopic and rooted in place, phytoplankton are microscopic unicells or colonies that float in the water. Historically, research into marine phytoplankton was
prompted by a desire to understand why fish stocks vary in abundance. Current
research is motivated more by the need to understand how phytoplankton affect
atmospheric CO2 and other climatically active gases through their roles in the
oceanic and global carbon and nutrient cycles.
The total biomass of phytoplankton is only about 1 % of that of terrestrial plants.
As such, phytoplankton are relatively inconspicuous, although their presence and
abundance can still be detected from changes in ocean color. Remarkably, despite
their low biomass, marine phytoplankton are as significant as forests and grasslands
to global photosynthetic CO2 fixation (Table 1). Current estimates indicate that
phytoplankton account for about 45 % of global net primary production.
Our understanding of the ecology of marine phytoplankton ecology has been
obtained using four complementary approaches: (i) oceanographic surveys and time
series, (ii) field-based perturbation experiments, (iii) laboratory culture experiments, and (iv) numerical models (Table 2). In many cases, phytoplankton are
investigated within the context of multidisciplinary programs addressing wider
issues in marine ecology or biogeochemistry. Oceanographic surveys document
the spatial distribution of phytoplankton at a particular time, whereas time series
document the seasonal and interannual changes of phytoplankton at a fixed location. Perturbation experiments are used to test hypotheses concerning the effects of
altering physical, chemical, or biological variables on phytoplankton ecology.
Laboratory culture experiments are used to study the physiological ecology of
phytoplankton. Numerical models are used to test our understanding of phytoplankton ecology.
This chapter describes how the approaches listed in Table 2 are used to gain
an understanding of primary production and phytoplankton ecology in the open
sea. The chapter starts by describing the main types of phytoplankton
from both taxonomic (section Phytoplankton Diversity) and functional
486
Table 1 Annual net primary production (NPP) of the biosphere. All values are in petagrams of C
(1 Pg 1015 g)
Marine
Phytoplanktona
Oligotrophic
Mesotrophic
Eutrophic
Subtotal
Coastal fringec
Microphytobenthos
Coral reef algae
Macroalgae
Sea grasses
Salt marsh
Mangrove forest
Subtotal
Chemosynthesis and anoxygenic
photosynthesisd
Total marine
NPP Terrestrial
Forestb
9.2
Tropical rain forests
34.8
Broadleaf deciduous forests
5.6
Mixed broadleaf and needleleaf forests
49.6
Needleleaf evergreen forests
Needleleaf deciduous forests
Subtotal
0.3 Grasslands, shrublands, and extreme
environmentsb
0.6
Savannas
2.6
Perennial grasslands
0.5
Broadleaf shrubs with bare soil
0.4
Tundra
1.1
Desert
5.5 Subtotal
0.4 Cultivationb
NPP
56.4
17.8
1.5
3.1
3.1
1.4
26.9
16.8
2.4
1.0
0.8
0.5
21.5
8.0
Phytoplankton Diversity
Phytoplankton cells range in size from about 0.5 m in diameter to >1 mm in
length: this is roughly the same relative difference in size as between a bumblebee
(about 1 cm) and a blue whale (up to 30 m) or between a tuft of grass (50 cm high)
and a redwood tree (100 m tall). The smallest phytoplankton cells have very simple
spherical or elliptical shapes. The largest can be highly ornate with elaborate
17
487
Methodology
Time-series stations
Examples
Natural
Bermuda Atlantic time series (http://bats.
community bios.edu/)
Hawaii ocean time series (http://hahana.
soest.hawaii.edu/hot/)
Transects
Natural
Atlantic Meridional Transect (http://amtcommunity uk.org/)
Remote sensing
Natural
Coastal zone color scanner (http://
community oceancolor.gsfc.nasa.gov/CZCS/)
SeaWiFS (http://oceancolor.gsfc.nasa.
gov/SeaWiFS/)
Aqua/MODIS (http://modis.gsfc.nasa.
gov/)
Field-based
Microcosms (0.120 Natural
Nutrient limitation (Mills et al. 2004);
perturbation
L)
community elevated CO2 and temperature (Feng
et al. 2009)
experiments
Ocean acidification experiments
Mesocosms (1100 Natural
m3, http://mesocosm. community (Riebesell et al. 2013)
eu/)
Eutrophication experiments (Romero
et al. 2012)
Iron fertilization experiments (Boyd
Open-water nutrient Natural
community et al. 2007)
additions (2575
km2)
Laboratory
Phenotypic
Algal
Light limitation (Falkowski et al. 1985);
culture
(physiological)
culture
nutrient limitation (Sunda et al. 2009)
experiments
acclimation
Genetic adaptation
Algal
Adaptation to high CO2 (Lohbeck
culture
et al. 2012)
Numerical models Simulation modeling of
Nutrient, phytoplankton, zooplankton,
biogeochemical and ecological
detritus (NPZD) (Fasham et al. 1990)
processes subject to physical
Plankton functional group (PFG)
forcing
(Le Quere et al. 2005)
Self-assembling (Follows et al. 2007)
spines and/or wings. Given this wide range of sizes, it is often convenient to
differentiate phytoplankton into size classes. One approach is to separate cells of
different sizes by sequentially passing a sample though a set of filters of decreasing
pore size. The most commonly employed pore sizes are 200, 20, 2, and 0.2 m.
Colonies and very large cells that are retained on 200 m sieves are referred to as
mesoplankton. Organisms that pass through a 200 m sieve but are retained on a
20 m pore filter are referred to as microplankton, and those that are retained by a
2 m pore filter but pass through a 20 m pore filter are referred to as nanoplankton.
The cells that pass through a filter with 2 m pores are referred to as picoplankton.
Traditionally, quantifying phytoplankton abundance in the sea has involved
collecting cells from seawater in a settling chamber or on a filter. Then, microscopy
is used to count and identify individuals. Phase contrast microscopy is often used
488
Fig. 1 Transmitted and fluorescence light micrographs of a centric marine diatom. The left side
shows the brightfield image of two Thalassiosira weissflogii cells in different orientations, as seen
using differential interference contrast. The right side shows a 3-dimensional reconstruction of the
fluorescence signals in the same cells that arise from chloroplasts, nuclei, and lipid droplets. Red
autofluorescence of chloroplast is shown in red, double-stranded DNA in blue (stained with
DAPI), and neutral lipids in green (stained with Nile Red) (Courtesy of Philippe Laissue and
Narin Chansawang (University of Essex))
for phytoplankton that have minimum linear dimensions of about 510 m. This
approach is limited since many phytoplankton are difficult to differentiate from
heterotrophic protozoa. In addition, many phytoplankton are considerably smaller
than 5 m in size and lack distinguishing morphological features. This problem can
be overcome using the fact that chlorophyll a a pigment found in all phytoplankton emits red light when illuminated. This chlorophyll autofluorescence allows
phytoplankton to be differentiated from heterotrophic organisms. Epifluorescence
microscopy allows cells that contain chlorophyll a to be visualized (Fig. 1), which
is useful for enumerating all but the very smallest phytoplankton.
Analytical flow cytometry (AFC) is now in routine use for enumerating phytoplankton. Like epifluorescence microscopy, AFC uses the red autofluorescence of
chlorophyll a to distinguish phytoplankton from protozoa and bacteria. Individual
cells are entrained in a narrow stream of fluid so that they can be passed one at a
time in front of a laser. The light that is scatted by the cells provides an indication of
their size and the red autofluorescence an indication of their pigment content. Some
flow cytometers have the capability to obtain images of individual cells. Highthroughput counting using flow cytometry has been routine for picoplankton since
the 1980s, and the upper limit to the size range of phytoplankton cells that can be
readily measured by AFC continues to increase.
The remainder of this section introduces some of the most important taxa of
marine phytoplankton (Table 3), starting with the smallest photosynthetic organisms, the picoplankton, and progressing upward through the size classes to the
nano- and microplankton.
17
489
Kingdom Division
Eubacteria Cyanobacteria
Typical
size
Cell
(m)
covering
<1
Organic
<3
310
Prochlorococcus
Synechococcus
Unicellular
diazotrophs
Trichodesmium
440
Eukaryota Alveolata (alveolates) Class Dinophyceae 52,000 Naked or
(dinoflagellates)
cellulose
plates
Division Chlorophyta Class
1040
Naked or
(green algae)
Chlorophyceae
cellulose
Class
140
Naked or
Prasinophyceae
organic
scales
Division Cryptophyta Class
620
Naked
(crytomonads)
Cryptophyceae
Division Haptophyta Class
520
Organic
Haptophyceae
or CaCO3
scales
Stramenopiles
Class
2200
Silica
(heterokonts)
Bacillariophyceae
frustule
(diatoms)
Class
815
Naked or
Chrysophyceae
scaled
(chrysophytes)
Class
35
25100
Dictyochophyceae
m silica
(silicoflagellates)
skeleton
Class
1.55
Naked or
Pelagophyceae
organic
(pelagophytes)
wall
Class
50100 Naked
Raphidophyceae
(raphidophytes)
Flagella
Absent
One transverse
and one anterior
flagella
0, 2, 4, or
8 smooth flagella
18 flagella
Two equal
flagella
Two flagella and
a haptonema
Two unequal
flagella (male
gametes only)
Two unequal
flagella
1 or 2
One hairy
forward and one
smooth trailing
One hairy
forward and one
smooth trailing
Picophytoplankton
Prochlorococcus, typically found in tropical and subtropical waters, are the
most numerous phytoplankton in the ocean (and are potentially the most
abundant photosynthetic cells on the planet!). They are often found at concentrations of more than 108 cells per liter of seawater. Despite this abundance,
Prochlorococcus was overlooked until the mid-1980s; because of its small size
(typically 0.50.8 m in diameter), it could not be differentiated from
490
Table 4 Contributions of different size classes and/or taxonomic groups to global phytoplankton
biomass and net primary production (NPP)
Contribution to
global biomassa, b
(Pg C)
0.21
0.10
0.44
Dominant taxa
Picoplankton Prochlorococcus
Synechococcus
Picoeukaryotes
(Prasinophytes, Pelagophytes,
Prymnesiophytes, Chrysophytes)
Nanoplankton Prymnesiophytes, Pelagophytes,
Cryptomonads
Microplankton Diatoms
0.51
Diazotrophs
Trichodesmium, unicellular
0.09
diazotrophs
Contribution to global
NPPc, d (Pg C/year)
11
20
15
0.41
Contribution of different groups to mean phytoplankton biomass for the ocean as a whole is from
information in the MAREDAT global synthesis (Buitenhuis et al. 2012)
b
Contributions of diazotrophs to biomass are the arithmetic mean given by Luo et al. (2012)
c
Contribution of different size fractions to global oceanic annual primary production is from Uitz
et al. (2010)
d
Contribution of diazotrophs to primary production is based on the arithmetic mean N2 fixation
rate of 140 Tg N/year given by Luo et al. (2012) using a C:N ratio of 6 gC/gN to convert N2
fixation to C fixation
17
491
Nanophytoplankton
The distinction between pico- and nanophytoplankton is in fact arbitrary since there
is a continuum of cells with sizes from about 0.8 m in diameter to >5 m in length.
Flagellated nanophytoplankton make a significant contribution to primary production. These small cells are often called microflagellates (small flagellates) even
though most are <10 m in length and so should rightly be called nanoflagellates.
Among the microflagellates, cryptomonads have been much less extensively investigated than other groups, although they can make an important contribution to
phytoplankton biomass in coastal waters. Cryptomonads employ phycobilins as
major light-harvesting pigments instead of chlorophyll a/b or chlorophyll a/c lightharvesting complexes found in other photosynthetic eukaryotes. Some
cryptomonads are heterotrophic, and many are mixotrophic, supplementing photosynthesis by ingesting bacteria or absorbing dissolved organic matter. The
haptophytes (or prymnesiophytes) are distinguished from other flagellates by
possessing a haptonema, a coiled flagellum-like structure that is located between
paired straight flagella. The haptonema may be used in prey capture or to escape
from predators. The coccolithophorids are a subset of haptophytes that are covered
in calcium carbonate plates called coccoliths. Coccolithophorids are major contributors to carbonate deposition in deep-sea sediments and, as such, affect both the pH
and alkalinity of the ocean. One of the most well-studied species of
coccolithophorids is Emiliania huxleyi, which sometimes blooms to such high
abundances that it imparts a chalky white color to the sea. The water masses that
contain these blooms can be seen from space!
Some unicellular cyanobacteria, also in the nanophytoplankton size class, are
capable of using nitrogen gas (N2) as a nitrogen source. Such organisms are termed
diazotrophs: from the prefix diazo which refers to two N atoms bonded together
and the suffix troph which means nourishment. Most bacteria and all eukaryotes are unable to use N2 because they lack the enzyme nitrogenase, which is
required to break the very strong N-to-N triple bond. Diazotrophs play a key role in
ensuring the continued fertility of the sea by fixing nitrogen into compounds that
can be used by other organisms. Their abundance and taxonomic composition is
most often assessed from the number of copies of nitrogenase genes (nif genes).
The potential importance of these unicellular photosynthetic diazotrophs to the N
budget of the ocean has only been recognized since the beginning of this century.
Microphytoplankton
Diatoms (class Bacillariophyceae) make the largest single contribution to global
oceanic net primary production, accounting for about 30 % of the total (Table 4).
Diatoms are characterized by being enclosed within a silica cell wall called a
frustule. The frustule is composed of two interlocking valves, much like a Petri
dish. The smallest diatoms are about 24 m in diameter and hence are a component of the nanophytoplankton (e.g., Minidiscus trioculatus), while the largest are
492
about 2 mm (e.g., Ethmodiscus rex). However, most diatom cells have diameters
between about 10 and 100 m, but their effective size is often increased by spines or
by forming colonies consisting of chains of cells. Although their silica frustule has
contributed to their evolutionary and ecological success, it can also be their Achilles
heel. Silicate, which is essential for building the frustule, can become depleted
before other nutrients, often bringing diatom blooms to an abrupt end, while other
phytoplankton that do not require silicate can continue to grow. Diatom blooms are
often followed by rapid export of organic matter from the illuminated surface
waters to the deep sea, accounting for as much of 50 % of the organic matter that
sinks to the deep sea.
Dinoflagellates (division Dinophyta) are unicellular organisms that have been
classified as algae by botanists and protozoa by zoologists. About half of all
dinoflagellate species are heterotrophic, with the remainder being photosynthetic
or mixotrophic. Dinoflagellates make a much smaller contribution to marine primary production than the diatoms, but nonetheless play important ecological roles
with significant economic impacts. They are motile and as such thrive under calm
conditions in stably stratified water columns. Some photosynthetic dinoflagellates
obtained their chloroplasts from the secondary endosymbiosis of red or green algae,
whereas others obtained chloroplasts from tertiary endosymbiosis of either a
crytomonad or haptophyte. Dinoflagellates often grow very slowly in nutrientpoor waters. Despite having low growth rates, dinoflagellates can form blooms by
employing effective defenses against grazers. The scales of armored dinoflagellates
are often shaped into spines or wings that provide a mechanical defense. Some
species are bioluminescent, emitting flashes of light when disturbed. These flashes
act as deterrents by making their zooplankton predators more conspicuous to fish.
Others produce toxins that affect mammals, birds, or fishes and are dangerous to
man when accumulated in seafood, such as oysters.
The filamentous cyanobacterium Trichodesmium is the most prominent diazotroph
in the sea. It is a major contributor to the input of fixed nitrogen to the tropical ocean,
particularly in the North Atlantic and Indian Oceans. Trichodesmium blooms are also
observed in waters around Australia and in the Red Sea. Trichodesmium is present as
individual filaments (trichomes) and also as large colonies of filaments. Surface
blooms, referred to as sea sawdust, rise to the surface under calm conditions.
When present at high abundance over large areas, these surface aggregations can be
detected using satellite ocean color sensors. Also contributing to N2 fixation in the sea
are diazotrophic heterocyst cyanobacteria, principally Richelia and Calothrix: these
are found in symbiotic association with some large diatoms.
Phaeocystis is a haptophyte genus that is widely distributed throughout the
ocean. Phaeocystis has received particular attention because it can form large
colonies consisting of hundreds or thousands of cells. The gel-like matrix of the
colonies is thought to store energy (polysaccharide) and nutrients (phosphate, iron),
whereas the skin of colonies is thought to prevent infection with pathogens and
present a mechanical barrier to zooplankton grazing. Phaeocystis blooms have been
reported in Arctic and Antarctic open ocean waters as well as nutrient-enriched
coastal waters.
17
493
494
17
495
sometimes called marine snow, also contribute to export production. These aggregates form when mucus nets produced by some zooplankton become clogged and
are discarded and/or when phytoplankton, bacteria, and detritus stick together to
form clumps. Fecal pellets and marine snow sink more rapidly when they include
dense mineral phases (e.g., calcium carbonate and biogenic silica) that are produced
by coccolithophorids and diatoms. Vertical migration of zooplankton and nekton,
which involves feeding near the sea surface at night and moving to deeper waters
during the day to avoid predators, can also contribute to export production. For
these reasons, export production is an emergent property that arises at the ecosystem level; the amount of export production cannot be inferred simply from the
properties of the components of the ecosystem, but relies on how these components
interact.
496
volume of large diatom cells that is occupied by the watery vacuole. For phytoplankton that lack vacuoles, such as dinoflagellates, the reduction in the cellular
density of carbon with increased volume is less pronounced.
Physiological traits including the cell-specific maximum nutrient uptake rate
and the affinity for nutrient assimilation also change markedly with increases of
cell volume. Much of the size dependence of these traits can be explained from
physical principles. For example, the rate at which nutrients diffuse to a cell should
be proportional to its radius. This leads to the expectation that the half saturation
constant for nutrient uptake should increase with (Vol)0.33. On the other hand, the
maximum rate at which nutrients can enter the cell is expected to be proportional
to the number of transporters on the cell surface, and this leads to the expectation
that the maximum rate of nutrient uptake will increase with (Vol)0.67. Such
constraints on physiology imposed by geometry and physics are most evident
when a very wide range of cell size is considered. However, there is considerable
variability that cannot be accounted for by cell size, and when working within a
restricted range of sizes, physiological sources of variability become increasingly
important.
17
497
Fig. 2 Shelf front transect. Data are collected through the water column across the transition
between shallow neritic shelf waters and deep oceanic waters as indicated by the water depth (a).
A marked decrease in the temperature of the sea surface (SST) is observed at the transition
between the shelf and ocean (b), termed the shelf break. Vertical cross sections of temperature
(c) and salinity (d) also display gradients at the shelf break, indicating the different water masses
on and off the shelf. The abundance of phytoplankton, as indicated from chlorophyll a (e) and the
availability of nutrients (f), also varies across the frontal transition
The average salt concentration of ocean waters is about 3.5 % (35 g of salt per kg
of seawater). Alterations occur when rainfall and river runoff decrease salinity by
dilution or when salinity increases due to evaporation or to ice formation. Salinity is
reported in practical salinity units (PSU), where 1 PSU is approximately equal to
1 g of salt per kg of seawater. Salinity is sufficiently uniform in the pelagic that it
has little direct influence on the physiology of phytoplankton over most of the open
ocean. However, salinity affects the density of seawater, which in turn influences
water motion and thus the supply of nutrients. Although salinity over most of the
sea surface ranges from about 30 to 38.5 PSU, marked deviations occur near coasts
where freshwater inputs are large and physical exchange with the open sea is
limited. Salinity varies from <0.5 to 20 PSU in estuaries and can be low in semienclosed waters such as the Baltic and Black Seas. At the other extreme, salinity
can reach very high values (>100 PSU) in brine-filled pockets of sea ice.
Oceanographic data is often presented as vertical profiles. These are plots of the
property of interest on the horizontal axis versus depth on the vertical axis. For
example, vertical profiles of temperature, salinity, and density for a location in the
subtropical North Atlantic Ocean are illustrated in Fig. 3 and corresponding profiles of
nutrients and dissolved oxygen in Fig. 4. Among the most conspicuous features in
498
Fig. 3 Vertical profiles of temperature, salinity, and density, expressed as sigmatheta density 1000 kg/m3 at the Bermuda Atlantic Time-Series Station. The permanent thermocline is evident at
depths below 500 m. Above this depth, temperature and density vary seasonally due to solar
heating and evaporation. BATS cruise 10106; 15 July 1997. Data provided by the U.S. National
Science Foundation funded Bermuda Atlantic Time-series Study Program: http://www.bios.edu/
research/hydrodata.html
Fig. 4 Vertical profiles of nitrate, phosphate, and dissolved oxygen at the Bermuda Atlantic
Time-Series Station. BATS cruises 10106 and 10107. 15 July 1997 and 1 August 1997. Data
provided by the U.S. National Science Foundation funded Bermuda Atlantic Time-series Study
Program: http://www.bios.edu/research/hydrodata.html
these profiles are the decreases of temperature and salinity and increase of density and
inorganic nutrients at depths between 500 and 1,000 m. These features persist throughout the year, and the depth zone from about 5001,000 m is called the permanent
thermocline. Also located in this depth zone is an oxygen minimum layer.
17
499
500
increasingly dominated by blue light in clear ocean waters and by green light in
coastal waters with higher amounts of phytoplankton. These changes in ocean color
are the basis for remote sensing of phytoplankton abundance using satellite-borne
sensors.
17
501
autumn and winter, density increases and the mixed layer deepens, eroding the
top of the seasonal pycnocline. Conversely, surface waters become lighter and
the mixed layer shoals when winter gives way to spring and summer.
Superimposed on the seasonal changes of mixed layer depths are diel changes
driven by the cycle of heating during the day and cooling at night. Deepening also
occurs when high winds increase vertical mixing by introducing turbulence at the
sea surface.
502
17
503
carbonic acid (Eq. 1), which in turn dissociates (breaks apart) to form carbonate
ions (Eq. 2) that in turn dissociate to form bicarbonate ions (Eq. 3).
CO2 H2 O $ H2 CO3
H2 CO3 $ H HCO3
H HCO3 $ 2 H CO3 2
Ocean Acidification
The ocean becomes slightly more acidic when CO2 dissolves in seawater because
protons (H+) are released during the reactions that form bicarbonate and carbonate.
There has been about a 30 % increase in the mean concentration of H+ in the
oceans surface waters during the past 250 years, and the rate of increase is
accelerating as more fossil fuel is burned. The mean pH of the surface waters has
decreased by 0.1 pH units over the past 250 years, from about pH 8.2 to pH 8.1,
and is projected to drop to as low as pH 7.9 by the end of the century. Although
these changes may seem small, they will be accompanied by marked decreases in
the concentration of carbonate ions and of the saturation state of carbonate
minerals with potentially dire consequences for marine organisms that produce
calcium carbonate shells (see section Anthropogenic Impacts on Marine
Phytoplankton).
Primary Production
The leaves of terrestrial vascular plants are essentially sugar factories that
produce sugars and starch during photosynthesis, with subsequent translocation
from mature leaves to the roots and actively growing tissues. For these plants,
504
This equation accounts not only for CO2 fixation but also for the assimilation of
nitrate and phosphate into organic matter. Somewhat paradoxically, this stoichiometry was obtained by examining the reverse process, namely, the decomposition
of organic matter in the deep sea, which leads to covariation in the concentrations of
nitrate, phosphate, dissolved inorganic carbon, and dissolved O2.
The ratio of O2 produced to CO2 fixed is called the photosynthetic quotient and is
designated PQ. Equation 4 gives a PQ of 150/106 1.45 mol O2 (mol CO2)1,
whereas this ratio is 1.0 mol O2 (mol CO2)1 for synthesis of sugars. The PQ is used
when comparing measurements of primary production based on O2 evolution (see
section The Photosynthesis-Irradiance Response Curve) with those based on CO2
fixation (see section Net and Gross Primary Production of Marine
Phytoplankton).
Oceanographers are concerned not only with gross primary production (GPP)
and net primary production (NPP) but also with net community production (NCP),
which takes into account the respiratory activity of heterotrophic organisms including bacteria, protozoa, and animals. The relationships among these different processes are conveniently summarized in the following equation:
NCP GPP RA RH NPP RH
17
505
XN
i1
Pi Zi
In this equation, N is the number of depth intervals, P(i) is the mean value of
NPP within the ith depth interval, and Z(i) is the width of that depth interval.
Throughout the first half of the twentieth century, oceanographers relied on the
measurement of O2 evolution as a proxy for CO2 fixation. This is because O2 has a
relatively low solubility in water, and very precise analytical methods for measuring O2 concentration have been available since the late nineteenth century. The
principle of the lightdark bottle method is simple. Briefly, O2 is produced by
photosynthesis in the light bottle, but is consumed by respiration in both the light
and dark bottles. NPP is obtained by measuring the increase of O2 concentration in
the light bottle, and respiration is obtained by measuring the decrease in of O2 in a
darkened bottle. GPP is then obtained from the sum of the increase of O2 in the light
and decrease in the dark.
In 1944, the American oceanographer Gordon Riley made the first estimate of
global oceanic primary production; this estimate was based on lightdark bottle O2
production determinations. Assuming a photosynthetic quotient of 1.45 CO2 fixed
506
Fig. 7 Primary production in the North Atlantic. Shown are vertical profiles of (a) gross O2
evolution (open circles) and 14C assimilation ( filled circles) during dawn-to-dusk incubations, (b)
chlorophyll a concentration at dawn, and (c) chlorophyll a-specific primary production rates (Data
are from Kiddon et al. 1995)
per O2 evolved, Rileys calculations give a value for GPP of the ocean as a whole of
87 56 Pg C per year (mean GPP of 234 151 g C m2 year1). Most of the
change that Riley observed was due to respiration in the dark bottle rather than NPP
in the light. The number of observations and their geographical range were very
limited and the error estimates associated with this calculation very large. In
addition, Riley employed incubations that lasted 3 days in order to obtain sufficiently large changes in O2 concentrations to be detected reliably. It was clear that a
more sensitive method for measuring primary production was needed.
A new method for measuring primary production was introduced to oceanography in
the 1950s by the Danish botanist and experimental biologist Einer Steemann-Nielsen.
Steemann-Nielsen developed the first protocols for using a radioactive isotope of
carbon, carbon-14 (14C), to measure marine primary production. The ease and sensitivity of the 14C method relative to the more cumbersome and less sensitive lightdark
bottle O2 method have allowed CO2 fixation to be measured routinely many thousands
of times. Current estimates of oceanic NPP (Table 1) rely on the accumulated database
of point 14C measurements of primary productivity that have been extrapolated to
the global scale using ocean color data (see section Remote Sensing of Primary
Production). The current estimate for NPP of the ocean based on extrapolation of the
14
C database is about 50 Pg C year1 (Table 5). This may be an underestimate as it only
accounts for the particulate carbon production, neglecting the carbon that is fixed into
dissolved organic matter, which can be a significant proportion of the total.
The lightdark bottle oxygen exchange method provides an unambiguous measurement of NPP, but may underestimate GPP because O2 uptake by phytoplankton is
often stimulated by light and therefore will not be accounted for by the consumption
that is measured in the darkened bottle. Two approaches used to obtain accurate
measurements of GPP rely on the stable oxygen isotope oxygen-18 (18O) and were
developed and applied in the 1980s. 18O accounts for only about 0.2 % of oxygen in
17
507
Mean Pg C year1
Range Pg C year1
21
12.8
9.9
2.6
0.33
0.45
47.1
15.530.9
9.117.9
6.915.1
1.14.9
0.021.2
0.280.73
3560
9.2
34.8
5.6
49.6
4.614.1
24.248.8
2.49.9
0.52
5.1
11.9
32
49.5
0.172.1
2.18.4
7.618.9
19.148.7
nature, but can be enriched to provide water and O2 that contain almost 100 % 18O.
Oxygen that contains two 18O atoms has a molecular weight of 36 (designated 36O2),
whereas oxygen that contains one 18O and one 16O atom has a molecular weight of
34 (designated 34O2) and oxygen that contains two 16O atoms has a molecular weight
of 32 (designated 32O2). Mass spectrometers are used to detect the amounts of O2 with
different masses. In the first approach, water that is labeled with 18O (i.e., H218O) is
added to a sample, and the production of 34O2 is measured (Eq. 7). In the second
approach, 36O2 is added to a sample, and its consumption is measured (Eq. 8). The
production of 34O2 from an illuminated sample that contains 18O-labeled water provides a direct measurement of the gross photosynthetic O2 evolution:
H2 18 O H2 16 O !34 O2 4 H
In contrast, a less direct approach to measuring GPP uses the consumption of 36O2
to obtain the rate of O2 consumption. In darkness, O2 is consumed by respiration of
organic matter.
Organic matter 36 O2 ! CO2 H2 O inorganic nutrients
508
the net O2 production in the light to obtain a value for GPP. GPP measured using
stable isotopes almost always exceeds estimates obtained from the lightdark bottle
technique.
One of the primary motivations for developing the 18O methods was to obtain
data that could be compared with 14C production to resolve whether the 14C method
yields results that are closer to NPP or GPP. During short incubations on the order
of minutes, it is anticipated that 14C production will be close to GPP because little of
the fixed 14C will have been respired back to CO2. However, as the duration of
incubations increases up to one day, more of the organic matter will be labeled with
14
C, and at least some of this will be respired back to CO2. Direct comparisons of
14
C production with gross O2 production have shown that the rate of CO2 fixation is
often about 50 % of the rate of gross O2 evolution. This difference is far larger than
can be explained by the photosynthetic quotient (see Eq. 4) and has been interpreted
to indicate that there may be a significant rate of light-dependent O2 uptake. Several
processes may account for this increase including the Mehler reaction, photorespiration, and light-stimulated mitochondrial respiration.
where (O2:Ar) is the change in the ratio O2-to-Ar during the time interval t.
A less direct way to estimate NCP is from information on the respiration of
organic matter in the deep waters below the euphotic zone. This is possible because
NCP is exported to the waters below the euphotic zone where more than 99 % is
respired, consuming O2 and releasing CO2. Since the 1950s, oceanographers have
been estimating the rate of respiration in the deep ocean from information on the
oxygen content and the residence time of water at different depths in the ocean.
In addition to 16O and 18O, there is a third stable isotope, 17O, which accounts for
only 0.04 % of the total oxygen. Geochemists have developed sensitive methods to
17
509
measure differences in the ratios 17O:16O and 18O:16O in O2, and these measurements can be used to estimate GPP without the need to incubate samples in bottles.
This triple isotope method relies on differences between the isotopic composition of
O2 added to the ocean by photosynthesis, O2 removed from the ocean by respiration, and O2 that dissolved into the ocean from the atmosphere. Photosynthesis
produces O2 that has the same isotopic composition as seawater. In contrast
respiration discriminates against the heavier isotopes and so increases the amounts
of 17O and 18O relative to 16O, with greater increases in 18O:16O than 17O:18O. The
isotopic composition of O2 in the atmosphere is not only affected by photosynthesis
and respiration but also by the exchange of oxygen between O2, O3, and CO2 in the
stratosphere. As a consequence of these processes, the atmosphere has a higher ratio
of 18O:17O than seawater. Taken together, these differences allow GPP to be
calculated from measurements of 17O:16O and 18O:16O.
510
observations are too few in number to calculate accurately the total primary
production of the ocean, let alone how primary production varies geographically
or how primary production changes through time. Fortunately, this problem can be
redressed by using information collected using satellite remote sensing.
Satellites provide three types of data that are used when inferring primary
production. These are chlorophyll a concentrations in the surface mixed layer, the
amount of solar radiation reaching the sea surface, and sea surface temperature.
Global distributions of chlorophyll a concentration are available from several
sensors including the coastal zone color scanner (CZCS) for the 1980s and more
recently the Sea-Viewing Wide Field-of-View Sensor (SeaWiFS) from 1997 to
2010 and MODIS (since 2002). Remote sensing of primary production relies on the
fact that the primary production of a water column is correlated with the mixed
layer chlorophyll a concentration. However, the correlation is not exact, which is
why oceanographers use additional information including location, solar irradiance,
and sea surface temperature in the calculations.
The earliest remote sensing estimates of primary production were made at local
and regional scales where empirical relationships had been established between
primary production and sea surface chlorophyll a. Extending the approach to ocean
basin and global scales required that more complex algorithms be developed,
several dozen of which have been devised. These differ in detail, but all rely on
mixed layer chlorophyll a being a robust index of the depth, chlorophyll a content,
and primary production of the euphotic zone. The basis of all algorithms is
calculation of NPP from the chlorophyll a concentration and chlorophyll a-specific
net photosynthesis rate:
NPP chl a Pchl
10
17
511
Fig. 8 Photosynthesislight
response curves for
Skeletonema costatum
acclimated to low light,
50 mol photons m2 s1
( filled circle), and high light,
1,200 mol photons m2 s1
(open circles). The same
observations of CO2 fixation
have been normalized to three
different indices of biomass:
to chlorophyll concentration
in panel a; to organic carbon
concentration in panel b; to
cell abundance in panel
c. Observations are for 14CO2
assimilation during 30-min
incubations and thus
approximate gross CO2
fixation (Data from the
experiments reported by
Anning et al. (2000))
512
Fig. 9 Dependence of (a) irradiance and (b) chlorophyll a-specific photosynthesis rate on optical
depth. Shown are the photosynthesis versus irradiance curves for Skeletonema costatum acclimated to low light and high light. The curves from Fig. 8 have been replotted versus optical depth
for a surface irradiance of 1,200 mol photons m2 s1. Optical depth is defined as ln(E(z)/E(0),
where E(z) is the irradiance at depth z and E(0) is the irradiance just below the sea surface. Optical
depths of 2.3 and 4.6 correspond to 10 % and 1 % of surface irradiance
11
A cells pigment content and composition, together with its size and shape,
combine to determine the value of achl; as a consequence, achl varies widely
between species, and it also varies with environmental conditions. The maximum
quantum yield (m) describes the greatest amount of photosynthesis that can be
17
513
achieved per unit photons absorbed. Cells actively alter m as they acclimate to
altered environmental conditions; for example, photoacclimation to high light
lowers m because photoprotective pigments are synthesized to dissipate more
absorbed light energy and hence transfer less to photosynthesis.
The maximum value of photosynthesis (Pmchl) is observed at irradiances where
light absorption no longer limits photosynthesis. What sets the value of Pmchl is
unclear; it may be limited by (i) the rate at which energy in the form of reductant
(NADPH) and ATP is delivered to the Calvin cycle, (ii) the rate at which CO2 is
incorporated into sugar phosphates by ribulose bisphosphate carboxylase, or (iii)
the rate at which the sugar phosphates produced by the Calvin cycle can be utilized.
The rate-limiting step varies between different species and/or in response to
different environmental conditions. More work to identify the mechanisms that
set Pmchl is essential given the importance of the light-saturated photosynthesis rate
in determining phytoplankton production and the possibility of CO2 limitation of
photosynthetic carbon fixation in some species.
Although Pchl is commonly reported by oceanographers and commonly
employed in bio-optical algorithms, it is a poor predictor of phytoplankton growth.
This is because there is taxonomic and phenotypic plasticity in the ratio of chlorophyll a-to-organic carbon. If, as is often the case, one wishes to know the specific
growth rate of phytoplankton, then in addition to measuring Pchl one must know the
ratio of chlorophyll a-to-carbon. This is because the three variables are related as
follows:
Pchl chl to C RA
12
In this equation, is the specific growth rate, Pchl is the chlorophyll a-specific
photosynthesis rate, [chl-to-C] is the ratio of chlorophyll a-to-carbon, and RA is the
respiration rate. Consequently, the characteristics of PE curves normalized to
chlorophyll a, organic carbon, or cell abundance differ (Fig. 8). Thus, care needs
to be taken when using information from these curves as quantitative traits.
Phytoplankton Ecology
The environmental factors that affect phytoplankton communities vary in time and
space in predictable and unpredictable ways. One particularly important predictable
pattern is in the seasonality of light and temperature in temperate and polar zones.
In these regions the total biomass of phytoplankton varies widely, with periods of
rapid proliferation in spring and autumn alternating with periods of decline. In
contrast, phytoplankton biomass is much more stable throughout the year in
subtropical and tropical waters that experience small changes in solar radiation
and where seasonal forcing by light and nutrient availability is much lower. Other
predictable patterns in phytoplankton community structure and primary production
are associated with the large-scale ocean circulation. Superimposed on these predictable patterns is randomness in solar radiation and nutrient availability due to
514
changing weather and currents. For example, sustained changes of wind speed and
direction in tropical waters can drive upwelling of nutrients to the surface, which in
turn drives changes in primary production.
At the broadest geographical scale, the oceans can be divided into four broad
domains (or biomes). These are:
High-latitude polar regions (where seasonal forcing is strongest)
Low-latitude (sub-)tropical regions (where seasonal forcing is weakest)
Intermediate mid-latitude regions
Coastal regions (where oceanic and atmospheric circulation patterns interact
strongly with the continents)
The location of physical oceanographic features, including pronounced horizontal gradients in temperature and salinity, is used to delineate different provinces
within each of the four domains.
17
515
Fig. 10 Archetypical seasonal production cycle in temperate waters. The spring phytoplankton
bloom occurs when solar radiation is sufficient to stabilize the water column and stimulate
phytoplankton growth. Nutrient depletion and/or grazing by zooplankton brings the bloom to an
end. Phytoplankton production in the mixed layer during the summer relies primarily on recycling
of nutrients. An autumn phytoplankton bloom occurs when the mixed layer deepens. This bloom
ends due to light limitation in deep mixed layer during winter
absorb heat and become more buoyant, and the mixed layer shoals. Consequently,
the average irradiance that phytoplankton experience increases, and a spring bloom
develops. The first quantitative explanation of the timing of the onset of the spring
bloom was developed by Harold Sverdrup and is referred to as critical depth theory.
The possibility that the onset of the spring phytoplankton bloom occurs as a
consequence of decreased grazing pressure exerted by zooplankton has recently
been proposed as an alternative to the traditional theory that the bloom starts simply
because the light environment becomes more favorable. The impact of zooplankton
on phytoplankton populations decreases rapidly when deep mixing dilutes the
abundances of both predators and prey. The reasoning behind this dilution hypothesis is that zooplankton will encounter phytoplankton much less frequently as both
populations decrease due to dilution. In the case where the water below the euphotic
zone is devoid of microorganisms, mixing equal volumes of deep water with
surface water will decrease the encounter frequency, and hence the mortality due
to zooplankton grazing, by a factor of four. Zooplankton populations, especially
protozoan populations, may decline as a consequence of food limitation, opening
up a window of opportunity for phytoplankton to escape being eaten by zooplankton when the water column begins to stabilize again.
As the spring bloom develops, much of the particulate matter sinks out of the
surface layer as fecal pellets or amorphous aggregates of particulate organic and
inorganic matter together with attached microorganisms. One explanation for the
end of the spring bloom is depletion of nutrients associated with this export. Not all
taxa are equally affected by nutrient limitation. In particular, the growth of diatoms
516
in the early stages of the spring depletes silicate, restricting further increases of
diatom populations. This typically occurs before nitrate is depleted, allowing other
phytoplankton taxa that do not require silicate, for example, the coccolithophorid
Emiliania huxleyi, the opportunity to bloom.
However, it is also possible that the bloom will peak before nutrients are
exhausted if the phytoplankton population is subjected to high rates of zooplankton
grazing or by outbreaks of viral disease. Whether the spring bloom is terminated by
nutrient limitation or high mortality, nutrients continue to be lost from the surface
mixed layer as organic particles continue to sink below the pycnocline. Subsequently, inorganic N becomes depleted, especially at lower latitudes (<40 N),
limiting primary production during the summer and causing the phytoplankton
community to shift to flagellate and picoplankton assemblages.
During summer, phytoplankton in the surface layer rely on the recycling of
nutrients, which can account for up to 8090 % of primary production at this time of
year. Consequently, primary production and heterotrophic consumption are tightly
coupled during this low nutrient period. A subsurface chlorophyll a maximum
(DCM) layer usually develops in the pycnocline at the interface between a
nutrient-limited zone at shallower depths and a light-limited zone below. The
phytoplankton in this layer intercept inorganic nutrients as they diffuse upwards
from below, and the DCM can make a significant contribution to primary production in summer.
As solar radiation declines in autumn, surface waters cool, increase in density,
and sink. This convective mixing erodes the pycnocline from above, transporting
nutrients and phytoplankton from the DCM into the surface mixed layer. This can
produce an autumn bloom, which eventually ends due to the light limitation as
autumn gives way to winter.
This description of the annual phytoplankton cycle emphasizes how limitation
by light and nutrients varies over the year. However, it has long been recognized
that phytoplankton populations increase in abundance much more slowly than
individual cells grow. The difference between the growth of individuals and the
growth of populations is due to mortality. The annual production cycle remains a
matter of active research because despite more than half a century of research,
debate continues on the relative importance of nutrient and/or light limitation of
individual growth versus mortality due to grazing and disease in controlling the
size, productivity, and species composition of phytoplankton communities.
17
517
variability in heat input is too small to generate enough cooling for convective
overturning to penetrate very far into the permanent thermocline. Consequently, the
convective input of nutrients to the surface is small. High solar radiation and the
limited extent of convective mixing throughout the year create conditions in which
phytoplankton growth and mortality remain tightly coupled resulting in low variability in phytoplankton biomass. The highly stratified regions of the subtropical
and tropical oceans are characterized by year round near-surface nutrient depletion
and relatively low uniform phytoplankton biomass (Fig. 11). Outside of the tropics,
the timing of the phytoplankton spring bloom varies not only with the seasonal
changes in the incident solar radiation but also with seasonal variability of mixed
layer depth.
The extent of convective mixing during winter is one of the main determinants
of the timing and magnitude of the bloom. Convective mixing increases at higher
518
Fig. 12 Experimental data from nutrient addition bioassay experiments conducted in (a) a
low-latitude N limited region of the subtropical Atlantic and (b) and higher latitudes in an
Fe-limited region. Differences in primary production measured by 14C incorporation are measured
in control samples and samples incubated with various concentrations of potentially limiting
nutrients (Replotted from Moore et al. 2006)
latitudes, for example, in the North Atlantic from about 150 m at 30 N to >800 m
at 60 N. Deeper convection leads to higher nutrient concentrations in surface
waters during winter. Deeper convection also decreases the extent to which the
growth of phytoplankton populations can be prevented by the grazing activity of
zooplankton. Together these two factors (higher winter nutrient concentrations and
lower topdown control of phytoplankton biomass by grazers) lead to more pronounced blooms at higher latitudes in the North Atlantic. For example, around 30 N
in the North Atlantic, rather than the bloom occurring in the spring, it occurs during
winter as nutrients are mixed into a well-lit surface layer. In contrast, in the regions
furthest to the north (>60 N), stratification is delayed, and the main phase of the
bloom occurs in summer. Year-to-year variability in weather (cloudiness and wind
speed), which influences both vertical mixing and the amount of solar radiation that
reaches the sea surface, can shift the timing of the bloom by several weeks.
Nutrient Limitation
Nutrient limitation has proven to be one of the more contentious issues in phytoplankton ecology. Up until the 1980s, geochemists were convinced that phosphorus
was the ultimate limiting nutrient in the sea, whereas biologists were equally
convinced that the key limiting nutrient was nitrogen. Geochemists maintained
that nitrogen could not be the limiting nutrient since N2 fixation would be used to
obtain nitrogen when other forms were exhausted. However, biologists had shown
from nutrient addition experiments that adding nitrate to samples stimulated phytoplankton growth but that adding phosphate on its own did not (Fig. 12) and so
17
519
concluded that nitrogen must be the limiting nutrient. The demonstration that iron
can be a limiting factor over large parts of the ocean in the 1980s and 1990s added a
new dimension to the debate between geochemists and biologists, but also helped to
reconcile their differences. It is currently thought that the input of iron to the ocean
limits that rate of N2 fixation, thus preventing the ocean as a whole from shifting
from nitrogen limitation to phosphorus limitation.
Nutrient limitation is often inferred from correlative studies that examine the
relationship between phytoplankton abundance or chlorophyll a concentration and
inorganic nutrient distributions over time (seasonal cycle) and/or in space (vertically in water column or horizontally along a transect). In these studies, low
concentrations of dissolved inorganic nutrients provide presumptive evidence of
nutrient limitation. However, limitation of growth rate is not proven because
recycling may be important and organic nutrients may be used. In addition,
covariation in the concentrations of limiting nutrients often precludes unambiguous
attribution to a single factor.
Presumptive evidence for limitation can be confirmed experimentally using
bioassays (Fig. 12). These involve collection of a large volume of water, which is
dispensed into a set of bottles to which the suspected limiting nutrients are added
alone and in combination, incubated under appropriate light and temperature
conditions and changes of biomass and other variables of interest are assessed.
Although widely used, such bottle experiments are not without their critics. In
particular, bioassay experiments perturb ecological processes that affect community structure such as predatorprey interactions or stimulation/inhibition of some
functional groups such as diazotrophs. To allow examination of ecosystem scale
responses to nutrient addition, oceanographers have turned to large-scale nutrient
fertilization experiments. Briefly, a patch of water about 10100 km2 in area is
enriched with the suspected limiting nutrient, and the increase of chlorophyll
a and declines in inorganic nutrients and CO2 are measured both inside and
outside the patch over a period of a few days to a few weeks. An inert tracer,
SF6, is added to account for advection and mixing. Such experiments have
confirmed that Fe is a limiting nutrient in oceanic regions where the macronutrients such as nitrate remain high, but where chlorophyll a concentrations remain
relatively low.
520
where surface currents converge. At the global scale, deep ocean waters containing
high concentrations of nitrate and phosphate are upwelled to the surface of the
Southern Ocean as a result of the westerly winds that circle the Antarctic continent.
These waters are advected to lower latitudes by surface currents or into the
permanent thermocline by subsurface currents. Subsequently, regional-scale
upwelling of cool, nutrient-rich water from the thermocline occurs in coastal
systems on the eastern boundaries of the low-latitude gyres and within the equatorial Pacific Ocean. These geographical patterns in resupply of deep ocean
nutrients to the surface drive similarly large-scale patterns in the extent and nature
of nutrient limitation, in phytoplankton distributions, and in pelagic ecology
(Fig. 11).
Nitrogen, phosphorus, and silicate availability tend to be low throughout the year
in the vast low-latitude subtropical and tropical oceanic regions, resulting in
persistently low phytoplankton standing stocks. Exceptions to this overall pattern
are observed within some coastal regions, where local upwelling can bring macronutrients to the surface. Also exceptional is the HNLC eastern equatorial Pacific
where strong upwelling brings macronutrients to the surface along the equator.
Away from these upwelling regions, the subtropical gyre regions which constitute
>50 % of the ocean surface, and hence >30 % of the whole Earth surface, are
highly oligotrophic. Dissolved inorganic forms of nitrogen (NO3, NO2, and
NH4+) are highly depleted in the subtropical gyre systems, and bioassay experiments have confirmed that nitrogen is the proximal limiting factor for primary
production in these systems.
As discussed previously, temperate and high-latitude North Atlantic waters are
characterized by a seasonal cycle; light availability restricts phytoplankton growth
in winter, while the lack of one or more nutrients contributes to the termination of
the spring bloom. However, the marked annual cycle of macronutrient (N, P, and
Si) concentrations and phytoplankton biomass that typifies the North Atlantic is
unusual when considered in the context of the global ocean. The annual cycle of
phytoplankton biomass and productivity is less pronounced in the other
mid-latitude and high-latitude systems, including the Southern Ocean and the
sub-Arctic North Pacific. Macronutrient concentrations remain high throughout
the year in these systems, while overall peaks in phytoplankton biomass
(or chlorophyll) are relatively low. Consequently, such regions are frequently
termed high-nitrate, low-chlorophyll (HNLC) regions. In these HNLC regions,
the concentrations of micronutrients, in particular Fe, are severely depleted.
The potential for Fe availability to play a major role in controlling phytoplankton production in these HNLC regions had been suspected for more than half a
century; however it wasnt until the 1980s that John Martin and colleagues
provided the first evidence in support of this hypothesis. Both bottle-enrichment
experiments (Fig. 12b) and experimental releases of dissolved Fe into the ocean
have demonstrated unequivocally that phytoplankton photosynthesis and growth
responds positively to the addition of Fe in the HNLC regions of the Southern
Ocean, the eastern equatorial Pacific, and the sub-Arctic North Pacific. Studies of
naturally iron-enriched coastal systems, for example, around sub-Antarctic
17
521
Islands, provide further support to this theory, and the Fe-limited status of the
HNLC systems is now widely accepted.
The existence of the HNLC systems can be understood by considering the
chemistry of dissolved Fe in seawater. Fe is highly insoluble and readily sticks to
particles in well-oxygenated seawater. Consequently, while inorganic nitrogen and
phosphorus are returned to the dissolved pool when organic matter decomposes,
iron remains attached to particles. These sink to the seabed, removing iron and
leaving behind an excess of dissolved nitrate and phosphate. Physical transport of
deep waters back to the surface supplies large amounts of the macronutrients nitrate
and phosphate, but very little Fe. It is therefore unsurprising that net growth of
phytoplankton depletes Fe before the macronutrients can be consumed, leading to
the development of Fe-limited systems. From this context, it is relevant to ask
Why do the macronutrients N and P ever become depleted to the point where they
become limiting?
The answer lies in considering the sources of Fe to the upper layer of the oceans.
The main inputs are from the Fe released from anoxic coastal sediments and from
dust generated in from arid regions and blown across the oceans by the wind. These
sources deliver large amounts of Fe to the lower latitudes of the North Atlantic,
which is one of the few Fe-replete ocean basins. In contrast, delivery of Fe to the
HNLC regions is insufficient to provide all of the Fe needed by phytoplankton to
fully utilize all macronutrients because the major HNLC regions are distant from
the largest dust sources. However, another important factor is that all of the HNLC
systems are characterized by high rates of deep mixing and/or wind-driven upwelling, which replenish macronutrients. Hence, a large annual supply of Fe would be
required to fully remove all the macronutrients from these systems. Conversely,
under conditions where the resupply of subsurface nutrients is slower, such as
within the stable highly stratified subtropical ocean gyres, dust and other fluxes of
Fe are sufficient to enable phytoplankton to fully utilize all the macronutrients.
The overall pattern of nutrient limitation at large scales can be summarized as
follows: iron is the limiting element in the upwelling dominated HNLC regions
which comprise around 3040 % of the oceans. Nitrogen is the limiting element
over most of the remainder of the ocean, dominated by the downwelling subtropical
gyres. Exceptional is the Mediterranean Sea in which N2 fixation and primary
production appear to be P limited, particularly in the eastern basin.
522
requirement for nutrients increases as the radius cubed (r3), and thus the growth rate
will decrease as the inverse of radius squared:
a S=r2
13
17
523
ammonium from organic molecules that are dissolved in seawater. Organisms that
use this strategy also express high-affinity nutrient transporters to insure uptake of
the ammonium and phosphate released by these enzymes. Other phytoplankton can
obtain nutrients by ingesting particles including bacteria and smaller phytoplankton
cells.
As previously discussed, nitrogen fixation is employed by Trichodesmium and
other diazotrophs to obtain nitrogen. However, diazotrophs require P and Fe in
addition to N, and these nutrients likely limit N2 fixation over large parts of the
ocean. High Trichodesmium abundances and high N2 fixation rates in the North
Atlantic Ocean occur downwind of the Sahara Desert and the semiarid Sahel
regions of Northern Africa due to deposition of wind-borne dust blown that
contains high amounts of Fe. Trichodesmium still requires P, which it can obtain
from hydrolysis of dissolved organic phosphorus compounds.
524
Table 6 Preindustrial, current, and projected future inputs of nitrogen, phosphorus, and silicate to
the ocean. Values are in Tg of N, P, or Si per year. The wide range of values between the studies
indicates the considerable uncertainty in these estimates
Nitrogen
N2 fixation
River discharge
Atmospheric
deposition
Phosphorus River discharge
Atmospheric
deposition
Silicate
River discharge
Gruber (2008);
Bennett et al. (2001)
Duce
et al. (2008)
Preindustrial 1990
135 50
135
50
30
80
20
6
50
20
8
22
1
1
1860
37
43
4148
1030 3896
5.9
6.6
8.48.5
142
144
136138
much lower effects. The assessment of the inhibition of primary production under
the ozone holes is complicated by difficulty in accounting for nonlinear effects of
UV-B and the interaction of UV-B with visible radiation in phytoplankton cells that
are subjected to vertical mixing.
The nutrient load to the ocean has increased dramatically over the past 300 years
as a result of population growth and intensification of farming practices. Some
estimates suggest that nitrogen and phosphorus inputs have increased by two to
three times above preindustrial levels, although there is considerable uncertainty as
calculations vary by about twofold (Table 6). Increased phosphorus and nitrogen
loading has not been evenly spread across the ocean. For example, loads to Chesapeake Bay have increased sixth- to eightfold and loads to the North Sea by about
10 times. Coincident with increased nutrient loading have been increases in the
incidence of harmful algal blooms (HABs) in coastal waters. Some HAB species
produce toxins, which can kill fish, shellfish, marine mammals, and/or seabirds. Algal
blooms can harm ecosystems in other ways. Persistent low oxygen (hypoxic)
conditions are found where O2 is depleted due to decomposition of organic matter
that has sunk from the surface to bottom waters and sediments. These dead zones
are found in the Gulf of Mexico under the Mississippi River plume, off the east costs
of Asia and North America and in coastal waters of Northern Europe. At the same
time that anthropogenic N and P inputs have increased, changes in the terrestrial and
freshwater nutrient cycling have led to a decrease in the inputs of silicate.
The increased nitrogen-to-silicate ratio that rivers deliver to coastal waters has
shifted the composition of phytoplankton communities away from diatoms and
toward flagellates, often decreasing the nutritional quality and palatability of the
phytoplankton. The input of nitrogen to the ocean from the atmosphere has also
increased due to emission of NO and NO2 accompanying combustion of
fossil fuels and emission of NH3 during the production and use of fertilizers.
17
525
Plumes of air polluted with these nitrogen compounds extend far downwind of major
population centers. Anthropogenically produced nitrogen compounds are being
deposited over almost all areas of the open ocean, with about 75 % of the nitrogen
deposition in regions that are nitrogen limited, and the input of anthropogenic
nitrogen into these regions is already approaching 50 % of the natural input due to
nitrogen fixation.
The average temperature of the atmosphere has increased by about 1 C in the
past 150 years. Most climate scientists attribute this to the increased concentrations of CO2 and other greenhouse gases in the atmosphere. The increase in air
temperature would have been dramatically larger were it not for the moderating
influence of the oceans over this time period. The oceans have absorbed about 40 %
of the CO2 released through burning fossil fuels and deforestation. This has slowed
the buildup of atmospheric CO2, which nonetheless is already over 1.4 times
higher than preindustrial levels. In addition, the oceans absorb a large amount of
heat that would otherwise warm the atmosphere; average sea surface temperature
(SST) has increased by about 1 C since 1880, and the interior of the ocean is also
warming.
Ocean warming has already affected the geographical distributions of plankton.
For example, there has been a well-documented northward shift in the distributions
of boreal and temperate copepod species in the North Atlantic Ocean. Ocean
warming will be accompanied by changes in ocean circulation and seasonal cycles
of stratification and mixing. The spatial extent of the subtropical gyres is expected
to expand, and the intensity of vertical mixing is likely to decrease. These regions
are characterized by year round or seasonally low macronutrient concentrations.
Consequently, increases in the area of these regions are likely to be accompanied by
a decline of oceanic net primary production. The flux of organic matter out of the
surface to the deep ocean (export production) is also likely to decline as stratification increases in the future. Although these processes will perturb the cycling of
carbon through the marine system, the feedbacks on atmospheric CO2 are not
simple to predict. They may be relatively minor, as the decreased export of organic
carbon should be balanced, in part, by a decrease in the return of CO2 and other
forms of inorganic carbon from the deep ocean to the surface.
Changes in atmospheric circulation and in the hydrological cycle, which are
accompanying global warming, are likely to affect the availability of iron to
phytoplankton. Changes in the areas of arid regions and changes in atmospheric
circulation will affect the amounts of iron delivered to different ocean basins by the
wind. Primary production will be stimulated if more iron is delivered to the ironlimited HNLC regions. Increased iron inputs to the nitrogen-limited subtropical
gyres may also stimulate primary production by reducing the extent to which iron
limits N2 fixation. Unfortunately, our understanding of the feedbacks in the climate
system is still too rudimentary to accurately predict how transport of atmospheric
dust to the oceans will change in a warming planet. Thus, it is also not possible to
predict the effect on marine phytoplankton. However, signals in the geological
record suggest that significant changes in oceanic primary production that have
occurred in the past were related to changes in Fe inputs.
526
The declining pH of the ocean due to invasion of the CO2 produced by mans
activities is called ocean acidification (see sections Dissolved Inorganic Carbon
and Ocean Acidification) is called ocean acidification (OA). Ocean acidification
is significantly altering the chemistry of seawater, including pH and CO2 (Fig. 13)
and calcium carbonate saturation state (Fig. 14). Critically, the current rate of pH
17
527
change is 100 times faster than the natural rates of pH change that have occurred in
the past. The potential influences of these changes on phytoplankton photosynthesis
and calcification have been investigated with laboratory monocultures and
mesocosm experiments.
Laboratory investigations on a small number of marine phytoplankton species
indicate that the response of growth rate to CO2 is most pronounced at CO2 levels
that are significantly lower than present-day values. Further increases of CO2 are
expected to have a negligible impact on the growth rate of most species. This lack
of response is likely due to the presence of carbon-concentrating mechanisms
(CCMs) that insure sufficient CO2 enters phytoplankton cells to meet the requirements for photosynthesis. Species in which growth rate increases in response to
elevated CO2 may lack CCMs or have inefficient CCMs. Some studies suggest that
growth of some picoplankton may be stimulated by elevated CO2 whereas others
suggest that it is microphytoplankton that benefit the most. However, even in these
cases, the effect of doubling CO2 from current levels is often small, typically less
than 10 %. Nonetheless, small differences in the response of growth rate to elevated
CO2 among species may still significantly affect phytoplankton community structure due to the cumulative effect of differences in exponential growth over many
generations. Unlike growth rates, which are largely unaffected by ocean acidification, the rate of calcium carbonate precipitation by coccolithophorids shows a
marked response. Although most studies show either no effect or a slight inhibition
of growth rate of coccolithophorids in elevated CO2, calcification usually declines
in response to OA, and the ratio of calcification to photosynthesis declines as a
consequence.
The insights from laboratory monoculture experiments do not allow assessment
of how OA affects species interactions, including competition for nutrients and
predatorprey dynamics. To address these issues, researchers have examined intact
plankton communities via experimental manipulations of closed systems (shipboard microcosms or in situ mesocosms) or observations of open systems made
along natural pH/pCO2 gradients. Open system observations take advantage of the
fact that low-pH seawater is found naturally, for example, upwelling of intermediate waters along the western North American continental margin and volcanic
CO2 vents in the Mediterranean and Indo-Pacific. These studies on intact communities have demonstrated that community structure responds to manipulation of pH
and pCO2. Nonetheless, results of these studies remain highly variable, thus limiting our ability to predict reliably the possible effects of increasing CO2 and OA on
phytoplankton productivity and ocean nutrient cycling.
Future Directions
Major unsettling of the earthatmosphereocean system including global
warming, ocean acidification, and cultural eutrophication is impacting marine
ecosystems. Currently, a predictive understanding of how these changes will affect
phytoplankton communities and productivity is lacking. Thus, a major focus for
528
ongoing and future research will be to document the changes in marine ecosystems
that are arising from anthropogenic activity and to develop a mechanistic understanding of why these changes are taking place. The goal is to obtain enough
knowledge to be able to make informed projections of the future state of marine
ecosystems and of the role of these ecosystems in global biogeochemistry. The
major questions include: How will phytoplankton species adapt to changing ocean
temperature and pH? How will phytoplankton communities be reorganized by the
responses to these changes? How will these changes in phytoplankton ecology
affect ocean biogeochemistry, for example, through release of climate reactive
trace gases? How will changes in phytoplankton influence higher trophic levels,
for example, impacting on fisheries yields, and how will overfishing affect phytoplankton ecology?
Satellite remote sensing allows us to measure how phytoplankton biomass varies
across the ocean. Calculating primary production from this information depends on
algorithms, which in the past have been developed from calibration against 14C
measurements. Ideally, these algorithms should instead be derived from first principles and then tested against the 14C measurements. Unfortunately, our understanding of the fundamental biological processes driving phytoplankton growth and
productivity (and how they are regulated by the environment) lags behind our
capability to measure biomass. Therefore, research needs to be undertaken to better
understand the ecophysiology of phytoplankton photosynthesis and the ecology and
evolution of phytoplankton communities.
To date, most studies of phytoplankton ecophysiology have tended to examine
one factor at a time, holding others constant. Although such studies can be useful
for gaining the most straightforward scientific insight, in an oceanic environment
where several factors naturally change simultaneously, it will be necessary to
conduct multifactorial investigations. However, because the number of experimental treatments that can potentially be investigated increases exponentially with the
number of different interacting factors under consideration, the design of such
studies needs to be informed by a clear understanding of how factors may covary
in both natural and anthropogenically perturbed systems.
The challenges of the multifaceted marine environment are particularly acute
when considering the biotic interactions that affect competition and succession.
Environmental change may simultaneously influence multiple trophic levels and
the interactions between them. In particular, sources of mortality remain relatively
underexplored when compared to the bottomup processes of resource limitation.
Mortality can arise from grazing by zooplankton and protozoa and/or by infection
by viruses and pathogenic bacteria. How these other components of the ecosystem
respond to climate change will no doubt be less predictable than those that will take
place in the physicalchemical environment.
Genomic, transcriptomic, and proteomic approaches have the potential to contribute to increasing our mechanistic understanding of the linkage between the
physiology of phytoplankton and their reciprocal interactions with the oceanic
environment. High-throughput sequencing is already revealing the high taxonomic
and metabolic diversity of marine phytoplankton alongside the complex integrated
17
529
References
Anning T, MacIntyre HL, Pratt SM, Sammes PJ, Gibb S, Geider RJ. Photoacclimation of the
marine diatom Skeletonema costatum. Limnol Oceanogr. 2000;45:180717.
Bennett EM, Carpenter SR, Caraco NF. Human impact on erodible phosphorus and eutrophication: a global perspective. BioScience. 2001;51:22734.
Boyd PW, et al. Mesoscale iron enrichment experiments 1993-2005: synthesis and future directions. Science. 2007;315:61217.
Buitenhuis, Li WKW, Vaulot D, Lomas MV, Landry MR, Partensky F, Karl DM, Ulloa O,
Campbell L, Jacquet S, Lantoine F, Chavez F, Macias D, Gosselin M, McManus
GB. Picophytoplankton biomass distribution in the global ocean. Earth Syst Sci Data.
2012;4:3746. doi:10.5194/essd-4-37-2012. www.earth-syst-sci-data.net/4/37/2012/
Carr M-E, et al. A comparison of global estimates of marine primary production from ocean color.
Deep-Sea Res. 2006;53(Pt II):74170.
Duarte CM, Cebrin J. The fate of marine autotrophic production. Limnol Oceanogr.
1996;41:175866.
Duce RA, et al. Impacts of atmospheric anthropogenic nitrogen on the open ocean. Science.
2008;320:8937.
Falkowski PG, Dubinsky Z, Wyman K. Growth irradiance relationships in phytoplankton. Limnol
Oceanogr. 1985;30:31121.
Fasham MJR, Ducklow HW, McKelvie SM (1990) A nitrogen based model of plankton dynamics
in the oceanic mixed layer. Journal of Marine Systems 48:591639.
Feng Y, et al. Effects of increased pCO2 and temperature on the North Atlantic spring bloom.
I. The phytoplankton community and biogeochemical response. Mar Ecol Prog Ser.
2009;388:1325.
Field CB, Behrenfeld MJ, Randerson JT, Falkowski P. Primary production of the biosphere:
integrating terrestrial and oceanic components. Science. 1998;281:23740.
Follows MJ, Dutkiewicz S, Grant S, Chisholm SW. Emergent biogeography of microbial communities in a model ocean. Science. 2007;315:18436.
530
Gruber N. The marine nitrogen cycle: overview and challenges. In: Capone DG, Bronk DA,
Mulholand MR, Carpenter EJ, editors. Nitrogen in the environment. 2nd ed. Amsterdam:
Elsevier; 2008. p. 150.
Jeffrey SW, Wright SW, Zapata M (2011) Microalgal classes and their signature pigments. In
Phytoplankton pigments characterization, chemotaxonomy and applications in oceanography
(Eds. Roy S, Llewellyn CA, Egeland ES, Johnsen G). Cambridge University Press. PP. 377.
Kiddon J, Bender ML, Marra J (1995) Production and respiration in the 1989 North Atlantic spring
bloom: An analysis of irradiance-dependent changes. Deep-Sea Res. 42:553576.
Le Quere C, et al. Ecosystem dynamics based on plankton functional types for global ocean
biogeochemistry models. Glob Change Biol. 2005;11:201640.
Lohbeck KT, Riebesell U, Reusch TBH. Adaptive evolution of a key phytoplankton species to
ocean acidification. Nat Geosci. 2012;5:16.
Luo YW, et al. Database of diazotrophs in global ocean: abundance, biomass and nitrogen fixation
rates. Earth Syst Sci Data. 2012;4:4773. doi:10.5194/essd-4-47-2012. www.earth-syst-scidata.net/4/47/2012/
Mills MM, Redame C, Davey M, LaRoche J, Geider RJ. Iron and phosphorus co-limit nitrogen
fixation in the eastern tropical North Atlantic. Nature. 2004;429:2924.
Moore CM, Mills MM, Milne A, Langlois R, Achterberg EP, Lochte K, LaRoche J, Geider
RJ. Iron limits primary productivity during spring bloom development in the central North
Atlantic. Glob Change Biol. 2006;12:62634.
Morel A. Optical modelling of the upper ocean in relation to its biogenic matter content (case-I
waters). J Geophys Res-Oceans. 1988;93:1074968.
Raven JA. Contributions of anoxygenic and oxygenic phototrophy and chemolithotrophy to
carbon and oxygen fluxes in aquatic environments. Aquat Microb Ecol. 2009;56:17792.
Riebesell U, Gattuso JP, Thingstad TH, Middelburg JJ. Arctic ocean acidification: pelagic
ecosystem and biogeochemical responses during a mesocosm study. Biogeosciences.
2013;10:561926.
Romero E, Peters F, Marrase C. Dynamic forcing of coastal plankton by nutrient imbalances and
match-mismatch between nutrients and turbulence. Mar Ecol Progr Ser. 2012;464:6887.
Seitzinger SP, Mayorga E, Bouwman AF, Kroeze C, Beusen AHW, Billen G, Van Drecht G,
Dumont E, Fekete BM, Garnier J, Harrison JA. Global river nutrient export: a scenario analysis
of past and future trends. Glob Biogeochem Cycles. 2010;24:GB0A08. doi:10.1029/
2009GB003587.
Sunda WG, Shertzer KW, Hardison DR. Ammonium uptake and growth models in marine
diatoms: Monod and Droop revisited. Mar Ecol Progr Ser. 2009;386:2941.
Uitz J, Claustre H, Gentili B, Stramski D. Phytoplankton class-specific primary production in the
worlds oceans: seasonal and interannual variability from satellite observations. Global
Biogeochem Cycles. 2010;24, GB3016. doi:10.1029/2009GB003680.
van Heuven SD, Lewis PE, Wallace DWR. MATLAB Program Developed for CO2 System
Calculations. ORNL/CDIAC-105b.Carbon Dioxide Information Analysis Center, Oak Ridge
National Laboratory, U.S.Department of Energy, Oak Ridge, Tennessee; 2009.
Further Reading
Arigo KR. Marine microorganisms and global nutrient cycles. Nature. 2005;437:34955.
Barton AD, Pershing AJ, Litchman E, Record NR, Edwards KF, Finkel ZV, Kirboe T, Ward
BA. The biogeography of marine plankton traits. Ecol Lett. 2013;16:52234.
Boyd PW, Strzepek R, Fu F, Hutchins DA. Environmental control of open-ocean phytoplankton
groups: now and in the future. Limnol Oceanogr. 2010;55:135376.
Cullen JJ, Boyd PW. Predicting and verifying the intended and unintended consequences of largescale ocean iron fertilization. Mar Ecol Prog Ser. 2008;364:295301.
17
531
Cullen JJ, Franks PJS, Karl DM, Longhurst A. Physical influences on marine ecosystem dynamics.
In: Robinson AR AR, McCarthy JJ, Rothschild BJ, editors. The sea, Biological-physical
interactions in the ocean, vol. 12. Boston: Harvard University Press; 2002. p. 297336.
Day TA, Neale PJ. Effects of UV-B radiation on terrestrial and aquatic primary producers. Annu
Rev Ecol Syst. 2002;33:37196.
Diaz RJ, Rosenberg R. Spreading dead zones and consequences for marine ecosystems. Science.
2008;321:9269.
Doney SC, Fabry VJ, Feely RA, Kleypas JA. Ocean acidification: the other CO2 problem. Ann Rev
Mar Sci. 2009;1:16992.
Falkowski PG, Barber RT, Smetacek V. Biogeochemical controls and feedbacks on ocean primary
production. Science. 1998;281:2006.
Falkowski PG, Katz M, Knoll AH, Quigg A, Raven JA, Schofiled O, Taylor JFR. The evolution of
modern eukaryotic phytoplankton. Science. 2004;305:35460.
Finkel ZV, Beardall J, Flynn KJ, Quigg A, Rees TAV, Raven JA. Phytoplankton in a changing
word: cell size and elemental stoichiometry. J Plankton Res. 2010;32:11937.
Geider RJ, et al. Primary productivity of planet earth: biological determinants and physical
constraints in terrestrial and aquatic habitats. Glob Change Biol. 2001;7:84982.
Katz ME, Finkel ZV, Grzebyk D, Knoll AH, Falkowski PG. Evolutionary trajectories and
biogeochemical impacts of marine eukaryotic phytoplankton. Annu Rev Ecol Evol Syst.
2004;35:52356.
Litchman E, Klausmier CA. Trait-based community ecology of phytoplankton. Annu Rev Ecol
Evol Syst. 2008;39:61539.
MacIntyre HL, Kana TM, Geider RJ. The effects of water motion on short term rates of
photosynthesis of marine phytoplankton. Trends Plant Sci. 2000;5:127.
Moore CM, et al. Processes and patterns of oceanic nutrient limitation. Nat Geosci.
2013;6:70110.
Rabalais NN, Turner RE, Justic D, Diaz RJ. Global change and eutrophication of coastal waters.
ICES J Mar Sci. 2009;66:152837.
Riebesell U, Tortell PD. Effects of ocean acidification on pelagic organisms and ecosystems. In:
Gattuso JP, Hansson L, editors. Ocean acidification. Oxford: Oxford University Press; 2011.
p. 99121.
Thomas MK, Kremer CT, Klausmeir CA, Litchman E. A global pattern of thermal adaptation in
marine phytoplankton. Science. 2012;338:10858.
18
Andrew D. B. Leakey
Contents
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Global Environmental Change from the Industrial Revolution to Today . . . . . . . . . . . . . . . . . . . . .
Greenhouse Effect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Greenhouse Gas Emissions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Temperature and Precipitation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Forecasts of Global Environmental Change in the Twenty-First Century . . . . . . . . . . . . . . . . . . . .
Plants as Pivot Points in the Global Carbon Cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Plants and Ecosystem Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Plant Responses to Elevated Carbon Dioxide (CO2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Photosynthetic Responses of C3 Species to Growth at Elevated [CO2] . . . . . . . . . . . . . . . . . . .
Respiration Responses of C3 Plants to Elevated [CO2] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Stomatal Conductance and Water Relations of C3 Plants Under Elevated [CO2] . . . . . . . .
Biomass and Seed Responses of C3 Plants to Elevated [CO2] . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Elevated [CO2] and Water Use Efficiency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Physiological Responses of C4 Species to Growth at Elevated [CO2] . . . . . . . . . . . . . . . . . . . . . . . .
Plant Responses to Temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Photosynthetic and Respiratory Responses to High Temperature . . . . . . . . . . . . . . . . . . . . . . . . .
Cellular Responses to High Temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Crop Reproductive and Yield Responses to High Temperature . . . . . . . . . . . . . . . . . . . . . . . . . . .
Carbon Cycling Responses to High Temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Plant Responses to Drought . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Stomatal, Photosynthetic, and Respiratory Responses to Drought . . . . . . . . . . . . . . . . . . . . . . . .
Plant Dehydration, Osmotic Adjustment, and Hydraulic Failure . . . . . . . . . . . . . . . . . . . . . . . . . .
Whole-Plant Physiological Plasticity and Adaptations to Drought . . . . . . . . . . . . . . . . . . . . . . . .
Crop Yield and NPP Responses to Drought . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
534
535
535
537
539
540
541
542
543
543
544
547
547
548
550
550
551
551
551
554
556
557
558
558
559
560
561
562
533
534
A.D.B. Leakey
562
562
563
565
566
567
570
571
Abstract
The Anthropocene is the period of Earth history since the Industrial Revolution and is defined by the impact of mankind on the environment.
Greenhouse gas concentrations, temperatures, precipitation, and atmospheric
pollutants have changed significantly from 1750 to today.
Climatic and environmental change will accelerate in the twenty-first century.
Plants act as pivot points in global biogeochemical cycles.
Plants provide many important ecosystem services including food production.
Elevated carbon dioxide (CO2) concentration enhances plant productivity.
Rising temperature stimulates plant productivity at high latitudes but impairs
plant productivity at many temperate and tropical latitudes.
Greater drought impairs plant productivity.
Elevated ozone (O3) concentration impairs plant productivity.
Crop plants can be adapted to future environmental change.
Future environmental change can be mitigated by appropriate management of
plants in agricultural and natural ecosystems.
Introduction
The term AnthropoceneAnthropocene emerged recently to describe the period of time
during which mankind has significantly impacted the function of the Earth system, i.e.,
biosphere, atmosphere, geosphere, ocean, and cryosphere. The use of the term is
intended to reflect the fact that since the start of the Industrial Revolution
(c. 17501850), humans have caused global environmental change comparable
with events that demark past geological epochs (e.g., the Holocene as the period of
12,000 years since the last ice age). The word Anthropocene has Greek roots, with
anthropo- meaning human and -cene meaning new. Human-induced changes in the
Earth system are occurring today at an accelerating pace and are anticipated to continue
for the foreseeable future. The resulting impacts on climate as well as ecosystem goods
and services are a growing challenge to human well-being. Secretary General of the
United Nations, Ban Ki Moon, in 2007 described climate change as the defining
challenge of our age. Recognition of this fact is a key driver of efforts to achieve
sustainable development, i.e., where current resource use meets human needs while
also preserving the environment to insure these needs can be met for future generations.
18
535
Plants mediate many key interactions between global environmental change and
humans. Plants play key roles in global biogeochemical cycles. For example, the
removal of carbon dioxide (CO2) from the atmosphere by plants via the process of
photosynthesis modifies greenhouse gas concentration in the atmosphere and the
greenhouse effect. In addition, plants play key roles in delivering the ecosystem
goods and services of food, fuel, fiber, forage, clean air, and clean water. Therefore,
any impacts of global environmental change on plants in natural or agricultural
ecosystems will influence human well-being.
This chapter takes a plant-centric view of the Anthropocene and aims to explain
(1) past and future global environmental change in the Anthropocene, (2) the role of
plants in global biogeochemical cycles and food security, (3) plant responses to
major elements of global environmental change (elevated CO2, temperature,
drought, elevated ground-level ozone), (4) adaptation of crop plants to global
environmental change, and (5) plant-based mitigation of global environmental
change.
536
A.D.B. Leakey
Fig. 1 Reproduced with permission from Ciais et al. (2013): Annual anthropogenic CO2 emissions and their partitioning among the atmosphere, land and ocean (PgC yr1) from 1750 to 2011.
(Top) Fossil fuel and cement CO2 emissions by category, estimated by the Carbon Dioxide
Information Analysis Center (CDIAC) based on UN energy statistics for fossil fuel combustion
18
537
Fig. 1 (continued) and US Geological Survey for cement production (Boden et al. 2011). (Bottom)
Fossil fuel and cement CO2 emissions as above. CO2 emissions from net land-use change, mainly
deforestation, are based on land cover change data and estimated for 17501850 from the average
of four models (Pongratz et al. 2009; Shevliakova et al. 2009; van Minnen et al. 2009; Zaehle
et al. 2011) before 1850 and from Houghton et al. (2012) after 1850 (see Table 6.2). The
atmospheric CO2 growth rate (term in light blue atmosphere from measurements in the figure)
prior to 1959 is based on a spline fit to ice core observations (Neftel et al. 1982; Friedli et al. 1986;
Etheridge et al. 1996) and a synthesis of atmospheric measurements from 1959 (Ballantyne
et al. 2012). The fit to ice core observations does not capture the large interannual variability in
atmospheric CO2 and is represented with a dashed line. The ocean CO2 sink prior to 1959 (term in
dark blue ocean from indirect observations and models in the figure) is from Khatiwala
et al. (2009) and from a combination of models and observations from 1959 from Le Quere
et al. (2013). The residual land sink (term in green in the figure) is computed from the residual of
the other terms and represents the sink of anthropogenic CO2 in natural land ecosystems. The
emissions and their partitioning only include the fluxes that have changed since 1750 and not the
natural CO2 fluxes (e.g., atmospheric CO2 uptake from weathering, outgassing of CO2 from lakes
and rivers, and outgassing of CO2 by the ocean from carbon delivered by rivers; see Figure 6.1)
between the atmosphere, land, and ocean reservoirs that existed before that time and still exist
today. The uncertainties in the various terms are discussed in the text and reported in Table 6.1 for
decadal mean values
538
A.D.B. Leakey
Fig. 2 Reproduced with permission from Ciais et al. (2013): Atmospheric CO2, CH4, and N2O
concentrations history over the industrial era (right) and from year 0 to the year 1750 (left),
determined from air enclosed in ice cores and firn air (color symbols) and from direct atmospheric
measurements (blue lines, measurements from the Cape Grim observatory) (MacFarling-Meure
et al. 2006)
18
539
higher temperatures, but is also highly reactive and degrades quickly. Therefore, O3
concentrations are highly variable in both time and space. Rising anthropogenic
emissions of methane and nitrogen oxides have caused ground-level O3 concentrations to increase from a preindustrial concentration of ~10 to ~40 ppb during
summer days in many parts of the world. In addition to being a greenhouse gas,
O3 is toxic to all life forms and significantly reduces the physiological performance
and productivity of all plants.
Unlike the other greenhouse gases described above, chlorofluorocarbons (CFCs)
and hydrochlorofluorocarbons (HCFCs) do not occur naturally and are solely the
product of industrial processes. They were used extensively during the 1900s for a
wide array of applications, including as refrigerants and aerosol propellants. However, CFCs and HCFCs were discovered to cause degradation of the high-level
(stratospheric) O3 layer that is responsible for absorbing harmful ultraviolet rays
from the sun. Consequently, a global ban on production of CFCs and HCFCs has
taken effect and substantially reduced emissions. As a consequence, concentrations
of CFCs and HCFCs peaked in the early 1990s, but the decline in concentrations is
slow due to the very long-lived nature of the molecules.
In summary, with the exception of CFCs, plants play important roles in the
emission of greenhouse gases and/or are directly influenced by changes in greenhouse gases that are toxic or important resources. In addition, greenhouse gases
indirectly impact plants by altering the climate.
540
A.D.B. Leakey
(i.e., greenhouse gas emissions) drivers of climate change correspond well with
measured data.
18
541
Fig. 3 Reproduced with permission from Collins et al. (2013): Time series of global annual mean
surface air temperature anomalies (relative to 19862005) from CMIP5 concentration-driven
experiments. Projections are shown for each RCP for the multi-model mean (solid lines) and the
5 % to 95 % range (1.64 standard deviation) across the distribution of individual models
(shading). Discontinuities at 2100 are due to different numbers of models performing the extension runs beyond the twenty-first century and have no physical meaning. Only one ensemble
member is used from each model and numbers in the figure indicate the number of different
models contributing to the different time periods. No ranges are given for the RCP6.0 projections
beyond 2100 as only two models are available
542
A.D.B. Leakey
Fig. 4 Reproduced with permission from Collins et al. (2013): Change in annual mean soil
moisture (mass of water in all phases in the uppermost 10 cm of the soil) (mm) relative to the
reference period 19862005 projected for 20812100 from the CMIP5 ensemble. Hatching
indicates regions where the multi-model mean change is less than one standard deviation of
internal variability. Stippling indicates regions where the multi-model mean change is greater
than two standard deviations of internal variability and where at least 90 % of models agree on the
sign of change (see Box 12.1). The number of CMIP5 models used is indicated in the upper right
corner of each panel
atmosphere by human fossil fuel burning and land-use change in 2009. While ~45
% of the CO2 emissions stayed in the atmosphere, ~30 % was absorbed by land
plants and ~25 % was absorbed by the oceans (Fig. 1; Ciais et al. 2013). These land
and ocean sinks for CO2 have significantly slowed the rate at which atmospheric
CO2 is rising and the climate is warming. However, the proportion of CO2 emissions absorbed by photosynthesis and stored on land or at sea is declining. Determining how future global environmental change will alter the performance of plants
and the control they exert on the global carbon cycle is therefore a scientific
priority. While these processes cannot be actively managed in natural ecosystems,
greater production of biofuels has the potential to increase carbon sequestration
while reducing fossil fuel use.
18
543
processes that drive ecosystem function. As described above, plants play a critical
role in the global carbon cycle. They also influence water cycling by acting as a
conduit for water to move from the soil to atmosphere through the process of
transpiration. Variation in plant cover or function that alters transpiration can
influence precipitation. For example, deforestation of the Amazon forest leads to
reduced transpiration, which in turn reduces convective rainfall and can intensify
drought. Plants also play important roles in global nutrient cycles; most prominently by interacting with microbes to perform nitrogen fixation. Through which
200 Mt of atmospheric nitrogen gas is converted each year across the globe into
chemical forms in the soil and ocean that are accessible to other organisms for
uptake. (2) Provisioning ecosystem services are actively harvested by us to meet
demand for natural resources including food, water, timber, and fiber. Approximately 1/8th of the plant biomass produced on the plant each year is harvested for
these purposes. Approximately 75 % of all calories consumed by humans come
directly or indirectly (via animal feed) from the four major crops of maize, wheat,
rice, and soybeans. (3) Regulating ecosystem services are processes in the Earth
system that control key physical and biological elements of our environment, e.g.,
climate regulation, flood regulation, disease regulation, and water purification. As
plants are the primary producers of all terrestrial ecosystems i.e., they synthesize
the carbon sources all animals and microbes subsequently use as energy sources
they are key to ecosystem stability and maintenance of regulating services. (4) Cultural ecosystem services reflect the aesthetic and spiritual values placed on nature as
well as the educational and recreational activities dependent on ecosystems. Plants
contribute to cultural ecosystem services as a result of mankinds emotional
response to time spent in a forest or a beautiful garden. Overall, plants strongly
influence human well-being through the services associated with both pristine,
natural ecosystems (e.g., tropical rain forests or arctic tundra), and highly managed
ecosystems (e.g., crop fields or urban landscapes). Consequently, the response of
plants to the elements of global environmental change in the Anthropocene (Fig. 5)
has and will continue to play a key role in determining the ultimate impacts on
human well-being.
544
A.D.B. Leakey
Ecosystem
Whole
plant
Leaf
Cell
Warming at
Elevated
Warming at
CO2
warm locations cool locations
Drought
Ozone
/
-/
Biomass
Leaf Area Index
Seed production
Defense
Senescence
-/
-/
-/
-/
-/
Photosynthesis
Carbohydrates
Respiration
Nutrient status
Water status
-/
-/
Enzyme stability
ROS
Membrane
stability
NPP
Shifts in
composition
of species or
genotypes
Fig. 5 Effects of global environmental change factors on plant processes at the ecosystem, wholeplant, leaf, and cellular scales. Arrows indicate direction of response (Modified from Ainsworth
et al. (2012))
and functional groups creates the possibility of important ecological and evolutionary consequences over a wide range of spatial and temporal scales. There are three
major photosynthetic types in higher plants: C3, C4, and CAM. C3 plants are the
most common and most sensitive to future, elevated [CO2]. C4 plants are less
common and less sensitive to elevated [CO2] but include some of the worlds most
important crops and weeds, as well as the grass species that dominate the worlds
tropical savannas. The response of CAM plants to elevated [CO2] has not been
studied extensively, in large part because they appear to be largely insensitive to
elevated [CO2]. Consequently, this section of the chapter focuses on the effects of
elevated [CO2] on C3 plants and then C4 plants.
18
545
A (mol m2 S1)
30
700
ci (mol mol1)
Fig. 6 The response of leaf CO2 uptake (A) to intercellular [CO2] (ci). Curves representing the
photosynthetic capacity of plants grown at ambient [CO2] (black solid line) and plants that have
undergone photosynthetic acclimation to long-term growth at elevated [CO2] (blue solid line) are
shown. The instantaneous stimulation of A when an ambient [CO2]-grown leaf experiences greater
internal CO2 supply after a shift from ambient [CO2] to elevated [CO2] is represented by the black
and orange dots. The stimulation of A when plants are grown long term in ambient [CO2] or
elevated [CO2] is represented by the black and blue dots. Dashed lines represent the decline in
[CO2] from outside to inside the leaf (supply function) associated with resistance to diffusion
through stomata
546
A.D.B. Leakey
18
547
548
A.D.B. Leakey
the leaf (stomatal conductance; gs). Variation of stomatal aperture provides plants
with dynamic control of the trade-off between carbon gain and water use. Growth at
elevated [CO2] leads to lower gs in almost all plants, with the exception of some
conifers and beech species. This is a direct and rapid response that appears not to be
modified by any acclimation of stomatal function after plants are grown at elevated
[CO2] for long periods of time.
The decrease in gs at elevated [CO2] acts to decrease transpiration per unit leaf area.
This in turn can decrease canopy-scale transpiration and overall crop water use at
elevated [CO2] compared to ambient [CO2]. However, the canopy-scale response is
usually more modest than the leaf-scale response due to aerodynamic conductances
between the leaf and the atmosphere and changes in leaf temperature that accompany
changes in gs. The relatively still air immediately next to a leaf (the leaf boundary
layer) becomes more humid as the leaf transpires. This process occurring on many
leaves collectively results in higher humidity within the plant canopy. This decreases
the gradient in humidity from the inside to the outside of the leaf that drives
transpiration. In dense, compact canopies where the air inside the canopy is rarely
mixed with the bulk atmosphere, transpiration can become significantly uncoupled
from stomatal conductance as a result. For example, gs of wheat is often >20 % less at
elevated [CO2] than ambient [CO2], but the resulting change in canopy evapotranspiration is <10 %. Two other factors also play a role in this response. First, total leaf
area per unit ground area (or Leaf Area Index, LAI) can be greater at elevated [CO2]
and offset the decrease in transpiration per unit leaf area. Second, the decrease in gs
and transpiration at elevated [CO2] results in less evaporative cooling of the canopy
and increases in leaf temperature. The internal air spaces of leaves are saturated with
water vapor (i.e., relative humidity 100%). Therefore, as leaf temperature rises
there is an exponential increase in the water vapor pressure of air inside the leaf, and
the gradient of water vapor pressure from inside the leaf to outside the leaf becomes
greater, driving greater transpiration for a given stomatal conductance.
Reduced canopy-scale transpiration at elevated [CO2] can ameliorate drought
stress by conserving soil moisture during drying events and delaying the onset of
stress. In addition, greater starting ci at elevated [CO2] and the nonlinear shape of
the A/ci curve mean that there is less inhibition of A by reduced CO2 supply (low ci)
when plants close their stomata in response to drought.
18
549
Growth at elevated [CO2] concentrations predicted to occur in the mid-twentyfirst century has been shown to stimulate the annual net biomass production
(defined as aboveground Net Primary Production; NPP) by approximately 20 %
over a broad range of temperate forest types (Norby et al. 2005; Norby and Zak
2011). Extra biomass has been shown to take the form of extra wood or greater fine
root production, depending on the forest type. This demonstrates the potential for
forests to absorb more CO2 from the atmosphere as atmospheric [CO2] rises and
slow the rate of climate change relative to anthropogenic carbon emissions. However, experiments fumigating entire forest canopies with elevated [CO2] in order to
test this possibility have been restricted to young, plantation forests in temperature
latitudes and relatively short periods of time (<15 years) compared to the life cycle
of most trees (tens to hundreds of years). This is significant because in some forest
experiments, elevated [CO2] stimulated NPP initially, but then the response diminished and stopped after approximately a decade. This pattern has been attributed to
a process called progressive nitrogen limitation. Progressive nitrogen limitation
occurs when stimulation of biomass production at elevated [CO2] locks up a large
fraction of the nitrogen in an ecosystem in inaccessible pools including wood and
soil organic matter. Over time, insufficient nitrogen is then available to support
stimulation of biomass production by greater photoassimilate availability. In some
forests exposed to elevated [CO2], faster release of nitrogen by microbial decomposition from soil organic matter occurs due to greater allocation of carbon from
trees to the microbes. This takes the form of greater exudation of carbon-rich
compounds from roots and greater carbon supply to mycorrhizae. When nitrogen
cycling is accelerated at elevated [CO2] in this manner it has prevented progressive
nitrogen limitation from occurring for a decade. However, it is not clear how long
such mechanisms can continue to operate. Mathematical modeling suggests that
progressive nitrogen limitation is likely in most temperate forests on multi-decadal
timescales. Experimental evidence for this is lacking, along with information on
how mature forests as well as tropical and boreal forests may respond to elevated
[CO2]. This is a significant source of error in projections of future carbon cycling
because of the large contribution of these particular forests to the terrestrial
carbon sink.
In C3 crops, greater biomass production is typically associated with greater seed
yield (Easterling et al. 2007; Tubiello et al. 2007; Leakey et al. 2009). Multiple
components of yield can contribute to this response, although an increase in the
number of seeds is usually more sensitive than an increase in the size of individual
seeds. Greater seed number can result from greater numbers of seeds per pod or
panicle or increases in the number of pods or panicles. On average, the major C3 crops
of wheat, rice, and soybean achieve ~15 % greater yield when grown in the field at
[CO2] expected for the mid-twenty-first century versus ambient [CO2] at the beginning of the century. However, there is significant variation around the mean driven by
genetic variation among crop varieties and environmental conditions. Genetic variation in crop yield response to elevated [CO2] could be exploited to identify key genes
that control sensitivity and provides one possible route to adapting crops for improved
performance in future growing conditions (Leakey and Lau 2012).
550
A.D.B. Leakey
18
551
552
A.D.B. Leakey
Toptimum
Rate of reaction
Temperature
30
l
tia
n
ote
Loss to photorespiration
al
actu
0
10
40
canopy temperature (C)
Fig. 8 Temperature response of potential and actual photosynthetic CO2 uptake (A), along with
loss of CO2 from photorespiration. Above the optimum temperature for photosynthesis, RuBisCO
activase and/or damage to photosynthetic membranes contributes to impairment of A
carboxylation. This results from a change in the specificity of Rubisco for oxygen
versus CO2, and a greater increase in the solubility of oxygen than CO2. As a
consequence, A (the net fixation of CO2 resulting from the balance of carboxylation
by RuBisCO and other processes releasing CO2, including photorespiration and
mitochondrial respiration in the light) increases initially as temperature rises, but
then reaches an optimum beyond which increases in photorespiration rate exceed
the rate of carboxylation (Fig. 8; Sage and Kubien 2007). In addition to this
18
553
554
A.D.B. Leakey
Fig. 9 Temperature
responses of dark respiration
where warm-grown (25 C)
plants are compared with
plants grown at hot
temperatures (35 C) and
displaying acclimation
responses. Dots depict
respiration rates at growth
temperatures and the arrow
represents the potential for
respiratory homeostasis
(Redrawn from Atkin
et al. (2005))
18
555
Heat stress causes dysfunction in the enzymes and bilayer lipid membranes that are
essential to cell function. This results in increased production of ROS from the
high-energy enzyme-driven reactions that take place in the thylakoid and mitochondrial membranes as part of photosynthesis and respiration, respectively. Further, ROS
are produced as a part of cellular stress signaling. The result is damage to a wide range
of complex molecules found in cells, including enzymes and other protein structures,
lipid membranes, and nucleic acids. High temperatures also increase water loss from
tissues and can result in tissue dehydration. Cellular dehydration can be an additional
cause of enzyme dysfunction, particularly for the large fraction of enzymes in the cell
which require sufficient water available to be solubilized.
A coordinated set of cellular responses occurs in response to sudden high
temperature exposure, which operate to counteract the negative effects of heat on
enzymes, lipid membranes, ROS, and dehydration (Mittler et al. 2004; Mittler and
Blumwald 2010). Collectively, these are referred to as a heat-shock response, and it
is characterized by transiently increased expression of genes encoding heat-shock
proteins. Heat-shock proteinsHeat shock proteins are found in all forms of life,
including archaea, bacteria, plants, and animals. The temperature at which a plant
normally grows influences the temperature at which expression of heat-shock proteins is induced. But, in general, heat-shock proteins in higher plants are induced by
temperatures greater than 3840 C. Heat-shock proteins function by binding to a
wide range of structurally unstable proteins in response to many stresses in addition
to heat. The binding of heat-shock proteins helps to prevent aggregation of denatured proteins, renature proteins that are denatured, stabilize proteins as they are
being translated from RNA, and modify proteins to allow membrane transport. Five
classes of heat-shock proteins have been identified, based on their molecular
weights (HSP 100, HSP 90, HSP 70, HSP 60, and small [sm]HSP). They are
found throughout the cell. The gene expression of heat-shock proteins is controlled
by transcriptional regulators called heat-shock factors. Expression of heat-shock
factors and heat-shock proteins is associated with signaling that upregulates antioxidant metabolism, osmotic regulation, and changes in lipid membrane structure.
For example, heat stress induces expression of cytosolic ascorbate peroxidase as
part of upregulating antioxidant metabolism. Protection against dehydration is also
induced by upregulation of pathways producing compatible solutes and osmolytes
that stabilize complex molecules and increase osmotic potential. These small
molecules include mannitol, proline, and glycinebetaine. Membrane stabilization
is achieved at high temperatures by increasing the fraction of lipids that are
saturated versus unsaturated. This raises the melting point in the same manner
that makes butter solid at room temperature when olive oil is liquid. As a result the
viscosity of the membrane can be maintained at levels that optimize its function as
an ion barrier and medium that supports proteins of diverse functions.
Acquired temperature stress tolerance and acquired thermotolerance are the
terms used to describe the phenomenon whereby a normally lethal temperature
can be survived as a result of being initially exposed to a high, but sublethal
temperature. For example, Arabidopsis seedlings grown at 22 C are killed by a
120-min exposure to 45 C. However, if the seedlings experience 38 C for 90 min
556
A.D.B. Leakey
18
557
involves abnormal pollen release from the anthers, resulting in less pollen and
unviable pollen reaching the stigma. This is important because a threshold of 1020
viable pollen grains must reach the stigma to ensure successful pollination. Genotypic variation in anther size and structure has been related to heat tolerance.
Yield losses to rising temperature have been observed over the last 3040 years
in some temperate and tropical regions and are anticipated to worsen for many
crops across much of the globe. A latitudinal gradient of crop response is anticipated (Easterling et al. 2007; Tubiello et al. 2007). Where current temperatures are
low at higher latitudes, warming will likely increase yields. This is partly due to
higher average temperatures during the growing season and a lengthening of the
growing season. However, these gains will be modest since many of these regions
are not currently intensively farmed. Agricultural intensification would not be
favored because the soils are often low in nutrients, but contain very high levels
of organic matter that would be oxidized and released by microbial respiration as
CO2 if the land is converted to agriculture. In lower, warmer latitudes, losses of crop
yield are expected to be greater because (1) many crops already grow near or above
optimal temperatures and (2) greater humidity reduces evaporative cooling of crop
canopies, resulting in greater plant tissue temperatures relative to air temperature.
The rice production area of tropical/subtropical Asia provides a case study of a
cropping system that will be strongly negatively impacted by global warming. Rice
yields in this region are currently being reduced by at least 30 % for every degree
of increase in night temperature during seed filling above a critical threshold
of 2223 C. In addition, high daytime temperatures are causing reproductive
failure of rice at local scales. Temperature increases associated with climate change
will immediately exacerbate the mechanisms currently driving yield loss, while
also potentially exceeding the temperature thresholds of additional physiological
processes that are important in determining yield. It appears that observed reductions in yield resulting from high nighttime temperatures can be explained, at least
in part, by greater respiratory loss of carbon. Meanwhile, current-day yield losses to
high daytime temperatures are most commonly ascribed to reproductive failure,
with inhibition of photosynthesis expected to cause further yield loss as daytime
temperatures rise in the future.
558
A.D.B. Leakey
attributed to result from warming that has triggered greater drought stress in
interaction with greater attacks from insect and microbial pests and pathogens.
This disturbance is again predicted to continue and be associated with a net loss of
CO2 from the biome. Finally, stem growth and NPP of tropical forests have been
shown to correlate negatively with annual average daily minimum temperature,
suggesting that warm nights lead to greater respiratory carbon losses. As a consequence of these trends across many regions of the world, models of the global
biogeochemical cycling indicate that global warming will act to lower NPP and
reduce absorption of fossil fuel emissions from human activities. In fact, some
models predict an amplification loop will occur in which warming leads to loss of
CO2 from ecosystems, especially the Amazon forest, which in turn accelerates
warming and drought, before ultimately leading to forest collapse. However, there
is considerable uncertainty in the resilience of forest ecosystems to such a response
and the precise tipping point of warming and drying that would be required to
trigger it. Fire plays an important role in forest mortality. As trees die from stress,
they increase the fuel load so that the heat and extent of fires are increased. This in
turn leads to greater canopy loss and, especially in tropical areas, reduced local
convectional water cycling. This in turn further exacerbates drought stress. Notably, the minimum area of a tropical forest fragment required for populations of
mammals and birds to be self-sustaining is similar to that necessary to support
local convectional water cycling. Therefore, efforts to conserve forest patches to
support biodiversity may incidentally allow local climatic regulation by the
ecosystem too.
18
559
of water to a plant does not meet the demand of the plant for water. Drought stress
can therefore occur as a product of soil drying and/or desiccating atmospheric
conditions of low humidity and/or heat. Drought occurs with varying frequency
and intensity in all biomes and is the primary limitation to crop yield as well as
ecosystem productivity globally.
560
A.D.B. Leakey
Fig. 10 Response of
stomatal conductance (gs) to
variation in leaf CO2 uptake
(A), humidity (h), and [CO2]
in C3 and C4 species
Ah
[CO2 ]
gs
gs = go + m.
s
cie
10
m
C3
sp
4
cies m
C 4 spe
Ah / [CO2]
18
561
osmosis. Many of these osmotically active molecules also act as compatible solutes,
which improve protein stability under dehydrating conditions. The osmotic adjustment mechanism can maintain turgor during mild to moderate stress, but is eventually
overwhelmed during intense drought. The extent to which cell water potential can
decrease until the turgor-loss point is reached depends on cell elasticity. Cells with
highly elastic walls contain more water at full turgor; hence, their volume can
decrease more, before the turgor-loss point is reached. The elasticity of the cell
walls depends on chemical interactions between the various cell-wall components.
In addition to loss of turgor, failure of water supply to meet demand can cause
hydraulic dysfunction in the form of xylem cavitation. This occurs when the tension of
water in xylem becomes sufficiently high that xylem sap vaporizes and dissolved air
forms a bubble in the xylem vessel. This halts water flow in the xylem and damages the
ability of the plant to transport water to the leaves, resulting in accelerated dehydration
and stress. Most plants can repair cavitated xylem, and some do so frequently, but
there is delay in return to full function. The manner in which stomata respond to water
shortages influences the likelihood of cavitation. Isohydric species (e.g., soybean and
sunflower) are those where a threshold water potential triggers stomatal closure to
minimize further transpiration. Anisohydric species (e.g., poplar and maize) are those
that are relatively insensitive to leaf water potential and whose rates of transpiration
are consistently higher. These modes of action represent the extremes of a continuum
of behavior. Isohydric species avoid extreme low water potentials and therefore xylem
cavitation, but are more likely to suffer from carbon starvation due to limited CO2
supply. Anisohydric species are more likely to experience hydraulic failure in the long
term, but maintain greater carbon balance under mild stress.
562
A.D.B. Leakey
Given the fundamental trade-off between carbon gain and water loss for plants, it
is not surprising that evolutionary adaptations to avoid drought stress are highly
diverse. They include strategies to maximize water uptake and water storage as well
as minimize water use. The archetypal drought adaptation is Crassulacean Acid
Metabolism (CAM) a type of photosynthesis found in cacti and other plant groups
found in desiccating environments. Plants with CAM photosynthesis circumvent
the trade-off between carbon gain and water use by opening stomata at night and
closing them during the day. At night, cooler temperatures mean that the gradient of
atmospheric humidity from inside the leaf to the atmosphere outside is much
smaller and rates of transpiration are relatively low. Meanwhile, CO2 is captured
at night and stored as malic acid in the vacuole. During the day, this CO2 is
rereleased and assimilated by RuBisCO. CAM photosynthesis is often accompanied
by other drought adaptations such as water storing trunks, stems, or leaves; deep tap
roots; thorns to eliminate tissue loss to herbivores; slow growth rates; or large
underground storage organs. Slow growth rates represent a conservative strategy of
stress tolerance by slow resource use and gain. Alternatively, some species living in
dry environments avoid stress by being annuals with very rapid life cycles and longlived seeds that remain in the soil until conditions are favorable for growth.
A similar strategy of stress avoidance is found in seasonally dry forests of tropical
and Mediterranean climates, which largely restrict growth to wet seasons.
18
563
along with the ecosystem goods and services they provide, is also negatively
impacted by current [O3]. It is important to note that ground-level O3 pollution is
a different environmental problem to the ozone hole, which is a reduction in [O3]
found high in the stratosphere that normally acts to filter out harmful UV rays from
the sun.
Most ground-level O3 forms from a series of reactions that are catalyzed by
sunlight between methane, volatile organic compounds (VOCs), and nitrous oxides
(NOx). All of these gases are predominantly released into the atmosphere by human
activities. There is significant spatial and temporal variation in [O3] because it is
highly reactive and forms or degrades quickly. Formation of O3 is favored by high
temperatures and sunlight. Therefore, a clear diel cycle in [O3] is typically observed
with low [O3] at night, rising [O3] from shortly after dawn until midafternoon and
then declining [O3] in the evening. Sources of air pollutants such as vehicle
exhausts and fossil fuel burning industries drive greater local ozone formation.
Prior to the industrial revolution, [O3] was less than 10 parts per billion (ppb) and
this provides an estimate of natural background [O3]. Today, daytime summer
[O3] regularly exceed 40 ppb in many parts of the Northern Hemisphere. However,
ground-level O3 pollution is not restricted to urban areas and significant plumes of
elevated [O3] air often form or move over rural areas. And, in cities, recently
formed O3 can react with NOx precursors in a futile cycle of synthesis and
degradation. This chapter mainly focuses on the responses of vegetation to O3
pollution. But, vegetation plays an important role as a sink for O3 from the
atmosphere and can therefore mediate significant land-atmosphere feedbacks. The
short lifetime of O3 in the atmosphere means that successful regulation of air
pollution could lead to relatively rapid reductions in ground-level [O3]. However,
clean air legislation is poorly enforced in many regions of the world and groundlevel [O3] is predicted to rise on average in the twenty-first century, with significant
increases projected for Asia and the Middle East. Some long-distance transport of
[O3] does occur, especially in the stratosphere, and inversion events can bring this
distributed source of O3 pollution to ground level.
564
A.D.B. Leakey
molecules that can damage important classes of complex molecules found in cells,
including enzymes and other protein structures, lipid membranes, and nucleic acids.
Cells rapidly sense elevated ROS levels in the apoplast and a complex signal
transduction network involving plant hormones, calcium ions, and protein phosphorylation cascades is activated. As a result, the expression of defense genes is
increased, leading to upregulation of antioxidant metabolism as well as cellular
repair processes.
Elevated [O3] decreases A across a wide range of species and environmental
conditions. Reductions in A at elevated [O3] are associated with reduced gene
expression, protein content, and activity of RuBisCO and other photosynthetic
enzymes. Lower A in turn reduces the pool sizes of sucrose and starch at elevated
[O3]. The decrease in carbon gain at elevated [O3] is often compounded by greater
rates of dark respiration. This may be due to the greater demand for energy from
antioxidant, defense, and repair processes induced by elevated [O3]. For example,
there is evidence for elevated [O3] stimulating production of apoplastic ascorbate,
flavonoids, volatile terpenoids, and epicuticular waxes. In addition, elevated [O3]
commonly accelerates leaf senescence, reducing the lifetime over which a leaf can
be contributing as a source of photoassimilates to the plant. The reduction in carbon
supply to other growing tissue from the range of responses described above
frequently leads to impaired root growth.
The decrease in A at elevated [O3] drives a feedback mechanism resulting in
lower gs as well. Furthermore, there is evidence that O3 exposure can have a
direct influence on stomatal function. This includes sluggish or insensitive
stomatal responses to other environmental stimuli, including abscisic acid. This
implies that plants grown at elevated [O3] may fail to close their stomata in
response to soil drying and exhaust soil moisture resources leading to greater
productivity and yield losses to drought. However, other studies have reported
that elevated [O3] diminishes stress under drought by decreasing stomatal conductance and reducing plant water use. The need for greater understanding of the
mechanistic basis for interactions between elevated [O3] and drought or temperature is a key knowledge gap. On the other hand, many studies have indicated
that elevated [CO2] protects plants from O3 damage by reducing flux into the
plant due to reduced gs and by providing greater photoassimilate to fuel defense
and repair responses.
It is important to note that a distinction is often drawn between plant responses to
long-term exposure to moderate [O3] (defined as chronic exposure of weeks to
months at <150 ppb) versus short-term exposure to high [O3] (defined as acute
exposure of minutes to hours at >150300 ppb). Chronic O3 damage of the type
described in the sections above is the most common scenario in the natural world
and is often not evident from rapid visual inspection of leaves. Acute exposures
have most commonly been applied in experimental settings. However, locations
with extreme air pollution do experience [O3] in the range that causes acute
damage. Acute damage is characterized by programmed cell death and significant
production of visible lesions on leaves.
18
565
566
A.D.B. Leakey
is globally responsible for $1426 billion in crop losses each year. This corresponds to significant crop losses in the major crops of soybean (616 %), wheat
(712 %), rice (34 %), and maize (35 %). There is significant potential for these
losses to grow over the twenty-first century, particularly in Asia and the
Middle East.
18
567
568
A.D.B. Leakey
fuel inputs to make that amount of fossil fuel (e.g., oil refining), plus the net balance
of carbon in the agricultural ecosystem. It is complex to determine carbon budgets for various fuel choices, and the discipline of life cycle analysis has emerged
to calculate the biological as well as socioeconomic budget factors. It is important
to recognize that biofuels have great potential as a strategy to mitigate rising [CO2],
but this potential will only be met if the correct crops are grown in appropriate
locations.
The two largest sources of biofuels produced today are maize (or corn) ethanol
in the USA and sugarcane ethanol in Brazil. Maize is an annual crop requiring
significant fossil fuel use in the production of fertilizers and pesticides as well as to
drive the mechanized equipment used for planting, fertilizing, and harvesting.
Sugarcane is a perennial crop that can regrow after harvesting each year and is
typically replanted every 67 years. Sugarcane is also grown with lower application of synthetic fertilizers because it is perennial with higher nitrogen use efficiency and because green fertilizers from harvest straw or cover crops are applied.
Both crops produce large quantities of biomass per unit land area. But, sugarcane is
grown in Brazil on marginal quality land not used for grain production, while
maize is grown in the USA on prime land. In the case of maize, this introduces
competition for land between production of biofuel feedstock and production of
maize for use in animal feed, food processing, and industrial applications. Ethanol
is produced by either: (1) fermenting the sucrose stored in sugarcane stalks or
(2) fermenting sugars produced by physical and chemical degradation of starch in
maize kernels. The necessity for fossil fuel inputs to produce sugars from maize
starch introduces inefficiency relative to using sucrose from sugarcane.
18
569
570
A.D.B. Leakey
and reduces the likelihood of unwanted land-use change from natural ecosystems to
agriculture.
Most intensively farmed agricultural soils have been losing carbon from soil
organic matter to the atmosphere over recent decades. This occurs in large part as a
result of tillage, where physical disturbance of the soil surface and incorporation of
crop residue by plows allow aerobic respiration by soil microbes to metabolize the
residue and soil organic matter and release it as CO2. This CO2 emission can be
substantially reduced by no-till or reduced-till practices. These involve the use of
alternative soil preparation and seeding equipment to sow the crop while maximizing the crop residue left on the soil surface and minimizing disturbance of the soil
surface. No-till or reduced-till practices have the added benefits of reducing CO2
emissions and costs associated with the number of times a tractor works a field, as
well as reducing erosion and increasing nutrient retention.
Forests are a major store of carbon in the global carbon cycle. Roughly 15 % of
current annual anthropogenic CO2 emissions come from deforestation. Reducing
deforestation and encouraging reforestation are therefore powerful potential mitigation strategies. However, these are challenging goals to achieve because most
forests are unmanaged, in remote locations and in developing tropical countries
where deforestation is profitable and not heavily regulated. Reducing Emissions
from Deforestation and Forest Degradation (REDD) is a scheme that places
monetary value on carbon stored in forests and, thereby, uses market-based
economics to incentivize reforestation. The idea is to provide developing countries
with financial incentives to reduce national deforestation rates below a baseline
determined from historical trends or a future projection. Countries that were able to
demonstrate reductions in CO2 emissions from deforestation would then be able to
sell carbon credits on an international carbon market. However, there are significant scientific and political challenges to implementing this plan. The greatest
scientific challenge is to find methods to assess how forest carbon stocks change
through time with deforestation and reforestation over vast land areas. Modeling
and remote sensing approaches where aircraft or satellites gather data on forest
structure and extent are being developed to address this need.
Future Directions
Key knowledge gaps in understanding the role of plants in future global environmental change, and societys response to it, include
Interactive effects of multiple factors of environmental change
Thresholds in plant responses to environmental change
Genetic variation in plant responses to environmental change in agricultural and
natural ecosystems
Ecological and evolutionary interactions that could amplify or negate the physiological responses to environmental change observed in individual plant
genotypes
18
571
References
Ainsworth EA, Rogers A. The response of photosynthesis and stomatal conductance to rising CO2:
mechanisms and environmental interactions. Plant Cell Environ. 2007;30:25870.
Ainsworth EA, Yendrek CR, Sitch S, Collins WJ, Emberson LD. The effects of tropospheric ozone
on net primary productivity and implications for climate change. In: Merchant SS, editor.
Annual review of plant biology, vol. 63. Palo Alto: Annual Reviews; 2012. p. 63761.
Allen CD, Macalady AK, Chenchouni H, Bachelet D, McDowell N, Vennetier M, Kitzberger T,
Rigling A, Breshears DD, Hogg EH, Gonzalez P, Fensham R, Zhang Z, Castro J, Demidova N,
Lim JH, Allard G, Running SW, Semerci A, Cobb N. A global overview of drought and heatinduced tree mortality reveals emerging climate change risks for forests. For Ecol Manage.
2010;259:66084.
Atkin OK, Bruhn D, Hurry VM, Tjoelker MG. The hot and the cold: unravelling the variable
response of plant respiration to temperature. Funct Plant Biol. 2005;32:87105.
Barnabas B, Jager K, Feher A. The effect of drought and heat stress on reproductive processes in
cereals. Plant Cell Environ. 2008;31:1138.
Bonan GB. Forests and climate change: forcings, feedbacks, and the climate benefits of forests.
Science. 2008;320:14449.
Chaves MM, Maroco JP, Pereira JS. Understanding plant responses to drought - from genes to the
whole plant. Funct Plant Biol. 2003;30:23964.
Ciais P, Sabine C, Bala G, Bopp L, Brovkin V, Canadell J, Chhabra A, DeFries R, Galloway J,
Heimann M, Jones C, Le Quere C, Myneni RB, Piao S, Thornton P. Carbon and other
biogeochemical cycles. In: Stocker TF, Qin D, Plattner G-K, Tignor M, Allen SK,
Boschung J, Nauels A, Xia Y, Bex V, Midgley PM, editors. Climate change 2013: the physical
science basis. Contribution of Working Group I to the fifth assessment report of the intergovernmental panel on climate change. Cambridge, UK/New York: Cambridge University Press;
2013.
Collins M, Knutti R, Arblaster J, Dufresne J-L, Fichefet T, Friedlingstein P, Gao X, Gutowski WJ,
Johns T, Krinner G, Shongwe M, Tebaldi C, Weaver AJ, Wehner M. Long-term climate
change: projections, commitments and irreversibility. In: Stocker TF, Qin D, Plattner G-K,
Tignor M, Allen SK, Boschung J, Nauels A, Xia Y, Bex V, Midgley PM, editors. Climate
change 2013: the physical science basis. Contribution of Working Group I to the fifth
assessment report of the intergovernmental panel on climate change. Cambridge, UK/New York: Cambridge University Press; 2013.
Easterling W, Aggarwal P, Batima P, Brander K, Erda L, Howden S, Kirilenko A, Morton J,
Soussana J-F, Schmidhuber J, Tubiello F. Food, fibre and forest products. In: Parry M,
Canziani O, Palutikof J, van der Linden PJ, Hanson CE, editors. Climate change 2007: impacts,
adaptation and vulnerability. Contribution of Working Group II to the fourth assessment report
of the intergovernmental panel on climate change. Cambridge, UK: Cambridge University
Press; 2007. p. 273313.
572
A.D.B. Leakey
Fuhrer J. Ozone risk for crops and pastures in present and future climates. Naturwissenschaften.
2009;96:17394.
Leakey ADB, Lau JA. Evolutionary context for understanding and manipulating plant responses to
past, present and future atmospheric CO2. Philos Trans R Soc B-Biol Sci. 2012;367:61329.
Leakey ADB, Ainsworth EA, Bernacchi CJ, Rogers A, Long SP, Ort DR. Elevated CO2 effects on
plant carbon, nitrogen, and water relations: six important lessons from FACE. J Exp Bot.
2009;60:285976.
McDowell N, Pockman WT, Allen CD, Breshears DD, Cobb N, Kolb T, Plaut J, Sperry J, West A,
Williams DG, Yepez EA. Mechanisms of plant survival and mortality during drought: why do
some plants survive while others succumb to drought? New Phytol. 2008;178:71939.
Mittler R, Blumwald E. Genetic engineering for modern agriculture: challenges and perspectives.
In: Merchant SBWROD, editor. Annual review of plant biology, vol. 61. Palo Alto: Annual
Reviews; 2010. p. 44362.
Mittler R, Vanderauwera S, Gollery M, Van Breusegem F. Reactive oxygen gene network of
plants. Trends Plant Sci. 2004;9:4908.
Norby RJ, Zak DR. Ecological lessons from free-Air CO2 enrichment (FACE) experiments. In:
Futuyma DJ, Shaffer HB, Simberloff D, editors. Annual review of ecology, evolution, and
systematics, vol. 42. Palo Alto: Annual Reviews; 2011. p. 181.
Norby RJ, DeLucia EH, Gielen B, Calfapietra C, Giardina CP, King JS, Ledford J, McCarthy HR,
Moore DJP, Ceulemans R, De Angelis P, Finzi AC, Karnosky DF, Kubiske ME, Lukac M,
Pregitzer KS, Scarascia-Mugnozza GE, Schlesinger WH, Oren R. Forest response to elevated
CO2 is conserved across a broad range of productivity. Proc Natl Acad Sci U S A.
2005;102:180526.
Sage RF, Kubien DS. The temperature response of C-3 and C-4 photosynthesis. Plant Cell
Environ. 2007;30:1086106.
Tubiello FN, Soussana JF, Howden SM. Crop and pasture response to climate change. Proc Natl
Acad Sci U S A. 2007;104:1968690.
Wilkinson S, Davies WJ. Drought, ozone, ABA and ethylene: new insights from cell to plant to
community. Plant Cell Environ. 2010;33:51025.
19
Contents
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Emissions to the Atmosphere from Plants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Biogenic VOC Emissions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Moving VOC from the Leaf into the Atmosphere . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Transport Versus Chemistry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
In- and Above-Canopy Turbulent Transport . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Top-of-the-Canopy Fluxes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Transport in the Atmospheric Boundary Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Chemistry in the Troposphere . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Gas-Phase Chemistry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Atmospheric Particles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Impacts on Air Quality and Climate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Climate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Air Quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
The Climate-Air Quality Conflict . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Future Directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
574
575
575
582
583
584
586
587
587
588
592
593
593
594
596
597
597
C. Wiedinmyer (*)
Atmospheric Chemistry Division, NCAR Earth System Laboratory, National Center for
Atmospheric Research, Boulder, CO, USA
e-mail: christin@ucar.edu
A. Steiner
Department of Atmospheric, Oceanic and Space Sciences, University of Michigan, Ann Arbor,
MI, USA
e-mail: alsteiner@umich.edu
K. Ashworth
Ecosystems-Atmosphere Interactions Group, Karlsruhe Institute of Technology, GarmischPartenkirchen, Germany
Department of Atmospheric, Oceanic and Space Sciences, University of Michigan, Ann Arbor,
MI, USA
e-mail: kirsti.ashworth@kit.edu
# Springer Science+Business Media New York 2014
R.K. Monson (ed.), Ecology and the Environment, The Plant Sciences 8,
DOI 10.1007/978-1-4614-7501-9_7
573
574
C. Wiedinmyer et al.
Abstract
Introduction
Have you ever walked through a forest and noticed that pine forest smell? What
you smell are trace gases released from the forest plants into the atmosphere. These
gases are known as biogenic emissions or emissions released to the atmosphere
from biological sources. Trace gases, such as volatile organic compounds (VOCs)
and the oxides of nitrogen (NOx), are emitted to the atmosphere from organisms
through a variety of biophysical and biochemical processes and can play an
important role in local, regional, and even global atmospheric chemistry and
climate.
In the mid-twentieth century, scientists began to recognize the importance of
biogenic emissions to the physical states and chemical processes of the atmosphere. Went (1960) presented evidence that plants emitted organic compounds to
the atmosphere, and further hypothesized that the blue haze observed in rural
regions, such as over the Blue Ridge Mountains in the eastern United States, is
the result of biogenically released compounds that have reacted and condensed to
form atmospheric particles. Rasmussen (1970, 1972) began to identify specific
organic compounds that were emitted from plant and other organism sources rather
than anthropogenic (human-made) sources. Since that time, advances in measurement technologies have enabled the detection and identification of hundreds of
chemical compounds that are emitted from vegetation to the atmosphere. Some of
these compounds are important to atmospheric chemistry, air quality, and climate
due to the magnitude of their emissions and/or their reactivity with respect to other
chemical species. For example, Chameides et al. (1988) provided the first quantitative study to show the importance of biogenic VOC emissions for the production
of ozone (aka photochemical smog) in the southeastern United States. Many
studies have shown that controls on anthropogenic sources of pollution may be
ineffective, or even counterproductive, unless biogenic VOC emissions are considered. Therefore, the understanding and quantification of biogenic emissions are
critical for the development of accurate models of the chemistry of our atmosphere, air quality, and climate. In this chapter, emissions of biogenic compounds,
19
575
their exchange between the biosphere and the atmosphere, and their impacts on
atmospheric chemistry and climate are explored.
576
C. Wiedinmyer et al.
Table 1 Annual global emissions of biogenic compounds (Adapted from Guenther et al. (2012))
Compound class
Isoprene
Monoterpenes
Sesquiterpenes (SQT)
Oxygenated VOC
Bidirectional VOC
Stress VOC
Other VOC
Compound
Isoprene
a-Pinene
t-b-Ocimene
b-Pinene
Limonene
Sabinene
Myrcene
3-carene
Camphene
Other monoterpenes
a-Farnesene
b-Caryophyllene
b-Farnesene
Other sesquiterpenes
2-3-2 methyl butanol (MBO)
Methanol
Acetone
Ethanol
Acetaldehyde
Formaldehyde
Acetic acid
Formic acid
Ethene
cis-3-hexenol
Other stress VOC
Propene
Butene
Other VOC
half of the total biogenic VOC emissions (Guenther et al. 2012). Other commonly
emitted compounds are monoterpenes (compounds containing 10 carbons, C10H16)
and sesquiterpenes (compounds containing 15 carbons). Biogenic VOC emissions
include oxygenated compounds, alkanes, alkenes, and acidic compounds (Table 1).
In general, the spatial distribution of the emissions closely follows the spatial
distribution of vegetation on the globe. Figure 1 shows the global land cover and
land use distributions as observed by satellite instruments from space. Biogenic
emissions are closely aligned with these types of global vegetation maps. The
specific types, as well as the quantity of volatile organic compounds produced by
ecosystems, are highly dependent on distributions of plant species and growth form.
For example, most oak trees (Quercus) emit isoprene at high rates; however, pine
Fig. 1 Global land cover and land use, as defined by the MODIS Land Cover-Type Product (http://modis.gsfc.nasa.gov/data/dataprod/dataproducts.php?
MOD_NUMBER12)
19
Plant Influences on Atmospheric Chemistry
577
578
C. Wiedinmyer et al.
trees (Pinus) do not emit isoprene, but they do emit monoterpenes. Isoprene can be
emitted in large quantities from areas with tropical forests and deciduous hardwood
forests. Monoterpene emissions are largely emitted from areas where boreal or
temperate coniferous species dominate the ecosystems. This is reflected in the maps
of biogenic VOC emissions shown in Fig. 2.
The mechanisms by which VOCs are produced and emitted also vary. Observations have shown that some biogenic VOCs, such as isoprene and some monoterpenes, are emitted as a function of environmental conditions that change on short
time scales, most importantly temperature and light. The time scale for the response
of these emissions is similar to that of photosynthesis, and in fact, these emissions
have been shown to be metabolically connected to photosynthesis through various
processes in the chloroplast. Other controls, such as leaf age and leaf area index
(LAI), can impact emissions but to a smaller extent. The reasons for the emissions
of biogenic VOCs vary and are in some cases not fully understood. For example,
several hypotheses exist to explain the ultimate aspects of natural selection that
have led to the evolution of isoprene emissions from leaves. One hypothesis is that
isoprene emission provides protection from elevated temperatures or from high
levels of atmospheric oxidants like ozone (Sharkey et al. 2008). Other compounds
are emitted as a response to insect attack or other abiotic or biotic stresses. These
stresses include light intensity, temperature, moisture availability, and exposure to
ozone pollution and insect attack. Monoterpene production in leaves and needles
has been attributed to protection from abiotic stresses, similar to the case for
isoprene, in some species, and to protection from insect herbivory in other species.
Plants may produce monoterpenes in different tissues and store them at different
levels, depending on these variable adaptive roles. In some leaves, monoterpenes
are produced in chloroplasts and not stored, rendering them susceptible to immediate leakage to the atmosphere; these compounds are thought to be most
effective at protecting leaves from abiotic stresses such as extreme heat, light,
and drought. In other leaves, particularly the needles of coniferous species, monoterpenes are produced in the cells of resin ducts and blisters and are stored as a
means of deterring insect consumption; these compounds leak more slowly to the
atmosphere and are thought to be most effective at protecting leaves from the biotic
stress of herbivory.
To estimate the quantity of emissions, particularly for atmospheric chemistry
and climate applications, biogenic VOC emissions are commonly represented by
Eq. 1:
Ei EFi i
where Ei is the emission of compound i (mass area1 time1), EFi is the potential
emission rate of compound i at a set of standard conditions, and i is an activity
factor that accounts for all environmental and phenological variables that control
the emissions. EFi is also known as an emission factor and its value can be a
function of a specific plant genus or ecosystem type. Table 2 shows the emission
factors of isoprene and some selected monoterpenes for several specific tree and
Fig. 2 Global annually averaged emission rate estimates (mmoles compound m2 h1) of several important biogenic VOC species for using the Model of
Emissions of Gases and Aerosols from Nature (MEGAN) v2.1 and the Community Land Model version 4 (Guenther et al. 2012). [SQT sesquiterpenes]. Note
the different scales for each figure
19
579
Isoprene
600
3,000
1
7,000
10,000
7,000
10,000
11,000
2,000
4,000
4,000
1,600
800
200
1
Limonene
100
100
130
80
80
80
80
80
60
100
60
0.7
0.7
0.7
0.7
3-carene
160
160
80
40
30
40
30
30
30
100
30
0.3
0.3
0.3
0.3
t--Ocimene
70
70
60
150
120
150
120
120
90
150
90
2
2
2
2
-Pinene
300
300
200
120
130
120
130
130
100
150
100
1.5
1.5
1.5
1.5
a-Pinene
500
500
510
600
400
600
400
400
200
300
200
2
2
2
2
Other
monoterpenes
180
180
170
150
150
150
150
150
110
200
110
5
5
5
5
-Caryophyllene
80
80
80
60
40
60
40
40
50
50
50
1
1
1
4
Table 2 Emission factors (mg compound m2 h1) of selected compounds from different plant functional types (Guenther et al. 2012)
Other
sesquiterpenes
120
120
120
120
100
120
100
100
100
100
100
2
2
2
2
580
C. Wiedinmyer et al.
19
581
Fig. 3 Photos of leaf enclosure measurements in the laboratory and in the field
The responses to various environmental and ecological conditions, or the individual gamma () values, are also dependent on the type of emitted VOC compound. Controls on isoprene emissions are dominated by leaf temperature and light
exposure. Isoprene is not emitted during the nighttime when it is dark.
582
C. Wiedinmyer et al.
19
583
Fig. 5 A schematic of the layers of the troposphere relevant for biogenic VOC (Adapted from
Arya (2001))
584
C. Wiedinmyer et al.
(e.g., isoprene and sesquiterpenes). The reactivity of each type of biogenic VOC is
defined in terms of the lifetime of the compound (), which refers to the average
length of time that a molecule will reside in the atmosphere before engaging in a
chemical reaction that changes its chemical structure. Atmospheric lifetimes reflect
the balance between compound emission rate and chemical reaction rate, which in
turn depends on factors such as the concentrations of reactants, temperature, and the
presence of catalytic surfaces that can reduce reaction activation energies. Generally, the local lifetime of isoprene ranges from a few hours to a few days (Fuentes
et al.2000), the lifetime of monoterpenes is on the order of minutes to hours, and
sesquiterpenes, more complex molecules with multiple double bonds, have a
shorter reactivity time of seconds to minutes. As a result, compounds with the
shortest lifetimes, such as sesquiterpenes, have the potential for reaction within the
canopy air space and thus may be unable to escape the canopy and enter the
planetary boundary layer.
One metric to estimate the relative importance of the reactivity to the atmospheric transport is the Damkohler number, representing the ratio of the chemical
lifetime of the compound to the transport time out of the forest canopy. If this ratio
is low, it indicates that chemical reaction times are much longer than transport
times, and most of the emitted species will be transported out of the canopy.
However, as this number approaches and exceeds unity, then the chemical reactions
occurring within the canopy are faster than the mean vertical canopy transport time
and the compound may not be emitted to the atmosphere above the canopy.
Additionally, a Damkohler number near or exceeding one also indicates that there
will likely be spatial and temporal inhomogeneities of biogenic VOC within the
forest canopy. These inhomogeneities in biogenic VOC concentrations, as well as
the concentrations of the radicals that drive chemical reactions, can effectively
lower reaction rates, a process known as segregation (Dlugi et al. 2010). Therefore,
understanding the relative roles of transport and atmospheric chemistry is important
for understanding fluxes out of the top of a forest canopy and will vary depending
on the biogenic VOC in question.
The Damkohler number varies as a function of canopy structure and meteorological conditions. For example, if we assume an average canopy residence time of
3 min, and a chemical lifetime of 84 min for isoprene (Table 3; isoprene + OH
reaction), the Damkohler number would be 0.04, indicating that most of the
isoprene will be transported to the surface layer. However, at nighttime when
canopy residence times lengthen (e.g., 10 min), a more reactive compound such
as terpinolene (a sesquiterpene with a chemical lifetime of 1 min with NO3;
Table 3) would yield a Damkohler number of 10, indicating that most nighttime
sesquiterpene emissions will react before leaving the canopy.
19
585
Table 3 Calculated atmospheric lifetimes (t) of selected biogenic VOCs with OH, NO3, and O3
(Rate constants from Warneck and Williams (2012)). The atmospheric concentrations of OH,
NO3, and O3 at which the lifetimes were calculated are provided at the bottom of the Table
O3
Compound
OH
NO3
Isoprene
1.4 h
48 min
1.3 days
a-Pinene
2.7 h
5 min
4.7 h
t-b-Ocimene
37 min
2 min
44 min
b-Pinene
1.9 h
13 min
1.1 days
Limonene
51 min
3 min
1.9 h
Sabinene
1.2 h
3 min
4.8 h
Myrcene
39 min
3 min
51 min
3-carene
1.6 h
4 min
11 h
Camphene
2.6 h
51 min
18 days
b-Phellandrene
50 min
4 min
8.4 h
Terpinolene
37 min
21 s
13 min
b-Caryophyllene
42 min
2 min
2 min
a-Humulene
28 min
1 min
2 min
Methanol
6 days
178 days
> 4.5 year
Atmospheric lifetimes based on the following concentrations (molec cm3):
[OH] 2.0 106
[NO3] 5.0 108
[O3] 7.0 1011
586
C. Wiedinmyer et al.
shear, where horizontal wind flows can be disrupted and create intermittent turbulent air motions that aid the transport of biogenic VOC. There is increasing
evidence that much of this turbulent transport occurs through the mechanism of
coherent wind structures (Finnigan 2000). Coherent wind structures are defined as
distinct patterns of turbulence that occur at regular intervals and are described by
two types of motion: (1) A burst or ejection of air from within the canopy to the
atmosphere (representing upward motion) and (2) a sweep of air that brings air
from the atmosphere into the forest canopy. These bursts and sweeps are due to
instabilities in the air flow caused by the large differences in horizontal wind speeds
near the top of the canopy. This can be visualized as a type of intermittent canopy
venting.
Coherent structures, such as the sweeps and bursts, occur on time scales of
seconds to minutes and are an important factor in the flux of biogenic emissions in
and out of a forest canopy. While the role of coherent structures on the transport of
biogenic VOC has yet to be quantified, results of studies on the transport of other
trace gases suggest that biogenic VOCs are likely to be carried along with coherent
structures and, depending on their chemical reactivity, vented to the atmosphere.
Therefore, identifying these structures and quantifying their contribution relative to
within-canopy reaction rate are key to understanding biosphere-atmosphere
exchange.
Top-of-the-Canopy Fluxes
The flux out of the top of the canopy into the planetary boundary layer represents
the mass flux of biogenic VOC to the atmosphere, which is the most important
emission metric for determining the role of biogenic VOC on atmospheric chemistry and climate. Biogenic VOC flux is defined as the mass of carbon
(or compound) per area per time and can be measured in the field with several
different techniques. Some studies have measured the fluxes of biogenic VOC at the
leaf or branch level, where a leaf or branch is enclosed in a chamber and the flux can
be quantified by measuring the flow and input and output concentrations (e.g.,
Fig. 3). These results must then be scaled with the biomass within the enclosure to
represent the full canopy.
In addition to branch enclosure methods, micrometeorological methods are
frequently employed to measure fluxes out of the canopy. Micrometeorological
methods use high time resolution measurements of wind speed, including the
turbulent and advective wind components, to estimate transport. The two most
commonly used micrometeorological methods for biogenic VOC flux estimation
are relaxed eddy accumulation (REA) and eddy covariance (EC). The REA
method collects air samples at the top of the canopy in updrafts and downdrafts
of the wind to determine a top-of-the-canopy flux. The EC method uses fastresponse time measurements (e.g., 110 measurements per second) to derive
fluxes as the statistical covariance between the turbulent wind speed and the
time-dependent variance in VOC concentration. The EC approach is similar to
19
587
techniques implemented to measure surface energy fluxes (Foken 2008). Typically, the REA method is used when high-response chemical sampling of the fast
fluctuations of biogenic VOC concentrations is unavailable. EC measurements of
top-of-the-canopy fluxes of biogenic VOC are becoming more common in field
sampling due to newer measurement techniques. The EC method is also advantageous to determine the role of coherent structure transport on the top-of-canopy
fluxes of biogenic VOC, as the fast-response time measurements can indicate when
coherent structures are present.
An additional metric often used to represent the fluxes of biogenic VOC out of
the forest canopy is the escape efficiency (Stroud et al. 2005). The escape efficiency
is defined as the fraction of the mass flux of biogenic VOC transported to the
atmospheric boundary layer as compared to the mass flux emitted from vegetation.
An escape efficiency of one therefore indicates that all biogenic VOC that is emitted
is mixed into the atmosphere. Stroud et al. (2005) show that this escape efficiency is
high (0.9) for less reactive species (e.g., isoprene and a-pinene) but low (0.3) for
-caryophyllene (a sesquiterpene). This method has been employed in models to
scale top-of-the-canopy flux estimates by removing the effect of in-canopy chemistry, which may reduce the source emissions of some very reactive biogenic VOCs.
588
C. Wiedinmyer et al.
down into CO2 and H2O, the rate at which this occurs varies widely among compounds,
and a diverse range of chemical by-products is also produced. Species with atmospheric
lifetimes of hours to days can be transported to other parts of a region or continent,
whereas VOCs with longer lifetimes become well-mixed in the free troposphere and
can be transported across global scales. The implications of such atmospheric transport
are discussed in the section Impacts on Air Quality and Climate.
While the precise reaction pathways of each emitted compound are determined by
their chemical structure, as well as the atmospheric concentrations of their reactants,
some generalizations can be made. The remainder of this section focuses on the
chemistry governing the production and loss of ozone and the formation of secondary
organic aerosols (trace components of the atmosphere that are both climatically active
compounds and air pollutants) in which biogenic VOCs play a major role. Integration
of the topic of VOC emissions from plants, as discussed above, with that of VOC
reactions in atmosphere, as discussed in the next section, provides the true nexus
required to understand how plants affect atmospheric chemistry.
Gas-Phase Chemistry
Generally, VOC emissions from plants are highly reactive, with atmospheric lifetimes
on the order of seconds to hours. Once released into the atmosphere, biogenic VOCs
react rapidly with atmospheric oxidants, primarily the hydroxyl (OH) and nitrate
(NO3) radicals, and also ozone (O3) molecules. (It is important to note that hydroxyl
and nitrate radicals are chemically different than hydroxide and nitrate ions. Free
radicals contain one or more unpaired valence shell electrons and are thus highly
reactive. It will be instructive to the student to explore the different chemical natures
of radicals and ions. For example, see suggested reading by Seinfeld and Pandis
(2006)). Table 3 shows the lifetimes of selected VOCs with typical atmospheric
concentrations of OH, NO3, and O3. The reactions of biogenic VOC with these species
produce secondary products that include O3, stable organic nitrate compounds that can
be transported for long distances, as well as low-volatility compounds that can
condense to form particles in the atmosphere. These particles (also called aerosols)
can remain suspended in the atmosphere for relatively long periods of time.
In the troposphere (Fig. 5), ozone (O3) is a pollutant and it can be harmful to human
health, plants, and other man-made materials. (Tropospheric ozone is different in its
ramifications for life on earth than stratospheric ozone. Stratospheric ozone protects the
DNA in cells from mutagenic ultraviolet radiation, whereas tropospheric ozone damages cells by causing oxidation of membranes, proteins, and nucleic acids.) Tropospheric ozone is also a strong greenhouse gas. Tropospheric ozone is produced primarily
through photochemically initiated reactions involving oxides of nitrogen (NOx) and
VOC, including biogenic VOC species. The downward transport of ozone from the
stratosphere to the troposphere is an additional source of tropospheric ozone, this source
is small in comparison to the rate of chemical production in the troposphere itself. The
main sink for tropospheric ozone is chemical loss, but there is also a significant flux to
the surface where it is lost by the process of dry deposition (Royal Society 2008).
19
589
Alternatively, peroxy radicals can react with nitrogen oxide (NO) to produce
stable molecules as shown in the general example Series B:
Series B :
ROO NO ! RO NO2
Thus, although the reaction chains are mostly initiated by the OH radical, the
rate of chemical production and loss of ozone is governed by the termination
590
C. Wiedinmyer et al.
reactions that are followed by the peroxy radicals formed during the second reaction
phase. This process is dependent on the concentration of NOx. In very low NOx
environments, such as remote parts of the Southern Hemisphere and Pacific Ocean,
the mutual termination reactions (Series A) predominate. As OH radicals are
formed in the first instance through the photolysis of O3, this sequence of reactions
results in a net loss of tropospheric ozone.
At the moderate-NOx levels encountered over rural areas across much of the
world, peroxy radical reactions with NO (Series B) predominate. Furthermore, the
NO2 produced undergoes photolysis and breaks down into NO and O*:
NO2 hn ! NO O
As the energetically excited oxygen atom can then react to form either new OH
radicals or, more importantly, O3 molecules (as shown in the initiation reactions
above), these regions are ozone producing. The rate of O3 production in such
regions increases with increasing concentrations of NOx but are relatively insensitive to changes in VOC emissions. Such regions are often described as NOxlimited or NOx-sensitive (Sillman 1999).
At even higher NOx concentrations, for example in urban areas in industrialized
or industrializing nations, the OH radical tends to react directly with NO2 to
produce nitric acid (HNO3) as shown below:
OH NO2 N2 ! HNO3 N2
When this reaction dominates the termination stage, insufficient O* atoms are
produced to outweigh the loss of O3 through photolysis, and the rate of O3
production declines. In such regions, an increase of hydrocarbons through VOC
emissions increases the sink for OH, reducing the rate of HNO3 formation
below the rate of NO2 photolysis. This results in an increased rate of O3 production, and these regions are often labeled VOC-limited or VOC-sensitive
(Sillman 1999).
If NOx concentrations rise further, a phenomenon known as NOx titration
occurs, and ozone concentrations fall as O3 reacts directly with NO to produce NO2
and O2 (Royal Society 2008):
O3 NO ! NO2 O2
As the NO2 produced from this reaction can subsequently reform NO and
O3 through photolysis, there is a further consequence to such high levels of NOx.
NOx titration leads to a rapid cycling of nitrogen and oxygen compounds and
this effectively allows the NOx to be transported away from the emission region
(i.e., polluted urban environment) to regions of lower background NOx levels,
which may result in the enhanced formation of O3 downwind from the original
NOx emissions (Sillman 1999).
19
591
As well as governing the rate of production and loss of tropospheric ozone, the
gas-phase reactions of biogenic VOCs play a key part in determining the atmospheric concentrations of a number of other gas-phase trace constituents of the
atmosphere. Biogenic VOCs act as a major sink, particularly over land, for the OH
radical, the atmospheres most powerful oxidant. Emissions of biogenic VOCs
thus mediate the oxidative capacity of the atmosphere, affecting the atmospheric
lifetime of other chemical species, such as methane (CH4). Methane is oxidized in
a similar set of reactions to those described above for non-methane VOCs. Thus,
methane and other VOCs compete for hydroxyl radicals in the free troposphere.
Simulations performed with atmospheric chemistry and transport computer
models have demonstrated that including biogenic emissions of isoprene alone
can increase the atmospheric lifetime of methane by up to 20 %, as compared to
model simulations without isoprene (Forster et al. 2007). This is a result of direct
competition for the OH radical; reactions with isoprene reduce the global OH
budget by around 8 % in such simulations. Inclusion of biogenic methanol
emissions results in similar impacts, though of lesser magnitude. Methanol is
not only less reactive than isoprene, with an atmospheric lifetime ranging from a
few days near the surface to a few weeks in the cold upper troposphere, but is also
emitted in smaller quantities. Nevertheless, such emissions are sufficient to reduce
the global average atmospheric concentration of the OH radical by around 2 %,
thus further increasing the atmospheric lifetime of methane. As methane is a
potent greenhouse gas, knowledge of the chemical reactions that affect its atmospheric lifetime and the ways in which the emissions of VOCs from plants can
affect the lifetime are important issues to understand and to include in models of
climate change.
Biogenic VOCs, and particularly isoprene, also play a key role in the distribution
of reactive nitrogen (i.e., nitrogen that is available in a form that will readily react
with other species rather than bound into long-lived stable virtually inert compounds such as N2O) in the atmosphere through the formation of organic nitrates,
and in particular peroxyacetyl nitrate (PAN). PAN is a relatively long-lived compound, with an atmospheric lifetime of several months in the cold free troposphere.
Vertical mixing lifts PAN from the boundary layer and lower troposphere, where it
is formed from reactions involving peroxy radicals (ROO) and NOx to the free
troposphere. Once there, its longevity allows it to be transported long distances
before it is broken down by either thermal decomposition or photolysis, rereleasing
reactive nitrogen. Thus, PAN acts to transport reactive nitrogen away from its
source to other regions of the world. For some remote regions, the reactive nitrogen
that is released from transported PAN (and other organic nitrates) can be the main
source of NOx. Atmospheric chemistry and transport model simulations show
significant PAN increases in the remote tropics due to isoprene oxidation when
biogenic isoprene emissions are included. The release of reactive nitrogen in such
regions, where isoprene emissions are high and background levels of NOx are low
(i.e., NOx-sensitive regions), can lead to enhanced ozone formation by shifting
the region from a low- to a moderate-NOx regime, as outlined previously.
592
C. Wiedinmyer et al.
Atmospheric Particles
In addition to the impact on ozone and other gas-phase constituents of the atmosphere described above, emissions of many VOCs from vegetation into the atmosphere affect the concentration of atmospheric particles or aerosols. The biosphere
is a source of aerosols both directly through the release of particles such as pollen,
plant detritus, bacteria, or spores, and indirectly as a result of the atmospheric
reactions of gaseous compounds. The former are referred to as biogenic Primary
Organic Aerosols (bPOA) and the latter as biogenic Secondary Organic Aerosols
(bSOA). While bPOA are generally thought to be larger in size and, therefore,
rapidly deposit back to the land or marine surface, bSOA are longer-lived,
impacting the atmosphere via both chemical and physical pathways. Their respective atmospheric lifetimes are again reflected in the distances over which they can
be transported and hence the impact they have on local and regional air quality and
global climate (section Chemistry in the Troposphere).
Although the gas-phase reactions of biogenic VOCs are initiated through reactions with atmospheric oxidants to form peroxy radicals that go on to produce
ozone, as outlined above, the products of these and subsequent reactions are often
oxygenated species of lower volatility than the parent VOC. At sufficiently low
volatility, these products can partition into the particle (or aerosol) phase, either
through direct nucleation or by condensation onto existing particles (see, e.g.,
Hallquist et al. (2009) and references therein).
Detailed analyses of the composition of atmospheric aerosol have shown that the
majority of their mass is biogenic in origin, even in highly polluted regions where
urban anthropogenic emissions are dominant. However, the series of gas-phase
reactions involved in SOA formation are complex and have not been fully elucidated for even the most common of VOCs. This is further complicated by the fact
that VOCs and their products can also undergo reactions in the aerosol phase and
participate in heterogeneous reactions (i.e., those that occur between compounds in
the aerosol and gas phases). Knowledge of the processes of aerosol phase and
heterogeneous chemistry and their controlling factors is even more limited than that
of gas-phase atmospheric reactions (Hallquist et al. 2009). Our lack of understanding is clearly demonstrated by the mismatch between the magnitude and spatial
distribution of SOA predicted by current theory and observations of aerosol concentration and composition (see, e.g., Spracklen et al. (2011)), although some of
this lack of agreement is undoubtedly the result of the need to reduce and simplify
the reactions included in most atmospheric chemistry models.
The biogenic VOCs that are emitted in the largest quantities, such as isoprene
and methanol, as well as their reaction products, have very low yields of
low-volatility condensable products and hence particles. In spite of their low yields,
the magnitude of their emissions suggests they do contribute substantially to the
total global SOA yield; but it is the longer-chained, and much more highly reactive
(those with atmospheric lifetimes of seconds to minutes), biogenic VOCs, such as
monoterpenes and sesquiterpenes, that are currently believed to have the highest
yields of condensable products. Despite their low emission rate, the total
19
593
Climate
Biogenic VOCs, in particular the terpenoids and other reactive species, have
atmospheric lifetimes that are too short to directly affect global climate. Longerlived species emitted from the terrestrial biosphere can be transported for long
distances before reacting or decomposing and may survive long enough in the free
troposphere to become well-mixed and ubiquitous in the atmosphere. However,
their radiative forcing or global warming potentials and therefore climate impact
are, as stated above, extremely low. The same is true of the organic gas-phase
reaction products from biogenic VOCs.
By contrast, tropospheric ozone (O3) is a potent greenhouse gas. Estimates of its
accumulated radiative forcing since preindustrial times place it third, behind only
carbon dioxide and methane, in terms of contribution to anthropogenic global
warming (see Fig. 1.1, Forster et al. (2007)). However, compared to both CO2
and CH4, O3 is short-lived, with an atmospheric lifetime ranging from a few days to
several weeks in the upper troposphere. Ozone is therefore less well-mixed through
the troposphere, and its climate impacts are regionally heterogeneous. As NOx
emissions in industrializing nations rise, it is to be expected that large areas of the
tropics will be transformed from low- to moderate-NOx regimes. This will result in
a considerable increase in O3 production from biogenic VOC reactions, likely to be
sufficient to affect the climate in these regions. Furthermore, the gas-phase
594
C. Wiedinmyer et al.
Air Quality
As highlighted previously in this section, biogenic VOCs play an important role in
the chemistry that produces tropospheric ozone. Ozone was first identified as a
primary component of smog, and therefore a key atmospheric pollutant, in the
1950s (Haagen-Smit 1950, 1952). Background levels (i.e., annual average concentrations at rural sites) of ground-level ozone have now reached around 3040
ppbv in the Northern Hemisphere and about 20 ppbv in the less-polluted Southern
Hemisphere. Peak hourly concentrations of ozone of over 100 ppbv are regularly
experienced during episodes of photochemical smog, with instantaneous concentrations over 400 ppbv recorded, caused by high temperatures and strong sunlight
accelerating the production of O3 from its precursors as well as promoting emissions of biogenic VOCs (Royal Society 2008).
Exposure to high levels of ozone has been shown to reduce lung function and
cause inflammation of the airways (WHO 2005), and epidemiological studies from
around the world have linked high ozone concentrations to increased cardiopulmonary mortality. For example, it has been estimated that around 22,000 deaths each
year are attributable to ozone in Europe alone. Current air quality guidelines
suggest a maximum daily ozone exposure limit of 50 ppbv (WHO 2005), although
legal limits vary between regions, with Europe, for example, setting an exposure
19
595
limit of 60 ppbv (EC 2002). Although high concentrations of ozone usually occur
with high temperatures, and often with high concentrations of other pollutants, e.g.,
NO2 and PM10 and PM2.5 (themselves subject to air quality control regulations),
meta-analyses of cardiopulmonary mortality data from epidemiological studies
around the world have shown that it is possible to eliminate the effects of these
confounders and deduce a concentration-response curve for the effects of ozone
alone. Such analyses indicate that there is an increase of 0.61.0 % in daily
mortality for every 10 ppbv increase in daily maximum ozone concentration
above a threshold of 35 ppbv, and this response is significant to at least the 95th
percentile. There is also growing evidence that long-term exposure to much lower
levels of ozone causes chronic damage to respiratory function (WHO 2005).
As well as human health effects, ozone causes oxidative damage to vegetation.
Ozone deposition onto vegetation surfaces leads to uptake through the stomata and
subsequent oxidative damage to plant cells and functions. Such damage reduces
photosynthesis, decreasing a plants ability to assimilate carbon and therefore
reducing productivity and crop yield (Sitch et al. 2007). Seed production and setting
are also affected, propagating the impact through successive generations. Field
studies of vegetation, particularly cash crops, have shown clear evidence of a strong
link between reduced yields and accumulated damage due to high ozone concentrations. In Europe, this damage is measured using a cumulative metric known as
AOT40, defined as the sum of hourly ozone concentrations (during daylight
hours when the stomata are open) above a threshold of 40 ppbv over the growing
season of the crop, usually a 3-month period that varies according to latitude and
crop type (CLRTAP 2004). More recently, it has been demonstrated that cellular
damage can occur at air concentrations below the threshold of 40 ppbv in some
instances, but that vegetation can conversely remain unaffected by concentrations
above this level. As oxidative damage is governed by the rate of uptake of
atmospheric ozone through the stomata (regulated by climate, soil moisture, atmospheric ozone concentrations, and plant growth stage), work is ongoing to develop
flux-based criteria for measuring likely damage and identify critical levels for these
metrics (see, e.g., CLRTAP (2010)).
It has been demonstrated that such concentration-based measures may not be the
best way to identify areas at high risk of ozone damage to vegetation. Within Europe,
for example, parts of Spain experience high ground-level ozone concentrations
during the growing season; however, ozone fluxes into plant cells are relatively
low as the stomata tend to be closed due to water stress during episodes of high
ozone. Conversely, the East of England has much lower atmospheric ozone concentrations, but plant cells there have high ozone uptake as the water stress is lower and
the stomata tend to remain open. Hence, although ozone damage to vegetation has
been widely observed, robust methods to quantify such damage lag behind those
developed for health impacts. This is in spite of the clear recognition of the economic
and societal implications of the loss of food production due to such damage.
Aerosols have a very obvious impact on air quality, reducing visibility and
creating visible haze (Went 1960). Particulate matter is also the biggest single
cause of air quality-related health effects, with over two million deaths worldwide
596
C. Wiedinmyer et al.
attributable to particles each year (WHO 2005). While the majority of these occur
in the developing world and are linked to indoor air pollution and cooking practices
(WHO 2005), it is an issue that affects all regions. For example, around 280,000
deaths in Europe are thought to be caused by atmospheric particulate matter, an
order of magnitude higher than those attributed to ground-level ozone. Air quality
guidelines (WHO 2005) set limits for daily and annual exposure to aerosol particles
with diameters of less than 10 m (of 50 g m3 and 20 g m3, respectively) and
2.5 m (of 25 g m3and 10 g m3, respectively). In general, the smaller the
particle, the more dangerous it is to the respiratory system as it is able to penetrate
further, with particles below around 1 g able to reach the lung surfaces.
Unlike ozone, there are no recommended exposure limits for vegetation. Indeed,
it has been speculated that the production of SOA is beneficial to vegetation as the
increase in particle concentrations and possibly cloud cover results in a higher
fraction of diffuse radiation relative to direct sunlight. Diffuse sunlight occurs as
the aerosols reflect and refract incoming radiation resulting in radiation reaching the
surface from all directions rather than solely from above. Shading of the lower
canopy by leaves in the upper canopy is reduced, and lower leaves receive more
radiation and are able to assimilate more CO2 through photosynthesis, resulting in a
higher overall productivity.
19
597
Future Directions
Constraining the quantity and environmental controls on biogenic emissions
Developing improved models to simulate biogenic emissions based on climatic
conditions
Understanding the role of biogenic emissions in the formation of aerosols in the
atmosphere
Understanding the interactions between urban and anthropogenic emissions with
biogenic emissions
Understanding the interaction of biogenic VOCs, atmospheric chemistry, and
climate in a changing world
References
Arya SP. Introduction to micrometeorology. San Diego: Academic; 2001. 420 pp.
Carlton AG, Pinder RW, Bhave PV, Pouliot GA. To what extent can biogenic SOA be controlled?
Environ Sci Technol. 2010;44(9):337680.
Chameides WL, Lindsay RW, Richardson J, Kiang CS. The role of biogenic hydrocarbons in
urban photochemical smog Atlanta as a case-study. Science. 1988;241(4872):14735.
doi:10.1126/science.3420404.
CLRTAP. Manual on methodologies and criteria for modelling and mapping critical loads and
levels and air pollution effects, risks and trends. Convention on Long-Range Transboundary
Air Pollution (CLRTAP). 2004. Available on-line from http://www.icp-mapping.org
CLRTAP. Manual of methodologies for modelling and mapping effects of air pollution. Convention on Long-Range Transboundary Air Pollution (CLRTAP). 2010. Available on-line from
http://icpvegetation.ceh.ac.uk
Dlugi R, Berger M, Zelger M, Hofzumahaus A, Siese M, Holland F, Wisthaler A, Grabmer W,
Hansel A, Woppmann R, Kramm G, Mollmann-Coers M, Knaps A. Turbulent exchange and
segregation of HOx radicals and volatile organic compounds above a deciduous forest. Atmos
Chem Phys. 2010;10(13):621535. doi:10.5194/acp-10-6215-2010.
EC. Directive 2002/3/EC relating to ozone in ambient air. Brussels: Commission of the European
Communities; 2002. Available on-line from http://ec.europa.eu/environment/air/legis.htm
Finnigan J. Turbulence in plant canopies. Annu Rev Fluid Mech. 2000;32:51971. doi:10.1146/
annurev.fluid.32.1.519.
Foken T. Micrometeorology. Berlin: Springer; 2008. 328 pp.
Forster P, Ramaswamy V, Artaxo P, Berntsen T, Betts R, Fahey DW, Haywood J, Lean J, Lowe
DC, Myhre G, Nganga J, Prinn R, Raga G, Schulz M, Van Dorland R. Changes in atmospheric
598
C. Wiedinmyer et al.
constituents and in radiative forcing. In climate change 2007: the physical science basis. In:
Solomon SD et al., editors. Contribution of working group I to the fourth assessment report of
the intergovernmental panel on climate change. Cambridge: Cambridge University Press;
2007.
Fuentes JD, Lerdau M, Atkinson R, Baldocchi D, Bottenheim JW, Ciccioli P, Lamb B, Geron C,
Gu L, Guenther A, Sharkey TD, Stockwell W. Biogenic hydrocarbons in the atmospheric
boundary layer: a review. Bull Am Meteorol Soc. 2000;81(7):153775. doi:10.1175/15200477(2000)081<1537:bhitab>2.3.co;2.
Guenther AB, Monson RK, Fall R. Isoprene and monoterpene emission rate variability observations with Eucalyptus and emission rate algorithm development. J Geophys Res Atmos.
1991;96(D6):10799808. doi:10.1029/91jd00960.
Guenther AB, Jiang X, Heald CL, Sakulyanontvittaya T, Duhl T, Emmons LK, Wang X. The
model of emissions of gases and aerosols from nature version 2.1 (MEGAN2.1): an extended
and updated framework for modeling biogenic emissions. Geosci Model Dev. 2012;5
(6):147192. doi:10.5194/gmd-5-1471-2012.
Haagen-Smit AJ. The air pollution problem in Los Angeles. Eng Sci. 1950;14(3):713.
Haagen-Smit AJ. Chemistry and physiology of Los Angeles smog. Ind Eng Chem Res.
1952;44:13426.
Hallquist M, Wenger JC, Baltensperger U, Rudich Y, Simpson D, Claeys M, Dommen J, Donahue
NM, George C, Goldstein AH, Hamilton JF, Herrmann H, Hoffmann T, Iinuma Y, Jang M,
Jenkin ME, Jimenez JL, Kiendler-Scharr A, Maenhaut W, McFiggans G, Mentel TF, Monod A,
Prevot ASH, Seinfeld JH, Surratt JD, Szmigielski R, Wildt J. The formation, properties and
impact of secondary organic aerosol: current and emerging issues. Atmos Chem Phys. 2009;9
(14):5155236.
Penner JE, Hegg D, Leaitch R. Unraveling the role of aerosols in climate change. Environ Sci
Technol. 2001;35(15):332A40. doi:10.1021/es0124414.
Rasmussen R. Isoprene: identified as a forest-type emissions to the atmosphere. Environ Sci
Technol. 1970;4:66771.
Rasmussen R. What do hydrocarbons from trees contribute to air pollution? J Air Pollut Control
Assoc. 1972;22(7):53743.
Royal Society. Ground-level ozone in the 21st century: future trends, impacts and policy implications. Fowler D, editor. Science policy report 15/08. London: The Royal Society; 2008.
Seinfeld JH, Pandis SN. Atmospheric chemistry and physics from air pollution to climate
change. 2nd ed. Wiley, New York; 2006.
Sharkey TD, Wiberley AE, Donohue AR. Isoprene emission from plants: why and how. Ann Bot.
2008;101(1):518. doi:10.1093/aob/mcm240.
Sillman S. The relation between ozone, NOx and hydrocarbons in urban and polluted rural
environments. Atmos Environ. 1999;33(12):182145. doi:10.1016/s1352-2310(98)00345-8.
Sitch S, Cox PM, Collins WJ, Huntingford C. Indirect radiative forcing of climate change through
ozone effects on the land-carbon sink. Nature. 2007;448(7155):7914. doi:10.1038/
nature06059.
Spracklen DV, Jimenez JL, Carslaw KS, Worsnop DR, Evans MJ, Mann GW, Zhang Q,
Canagaratna MR, Allan J, Coe H, McFiggans G, Rap A, Forster P. Aerosol mass spectrometer
constraint on the global secondary organic aerosol budget. Atmos Chem Phys. 2011;11
(23):1210936. doi:10.5194/acp-11-12109-2011.
Stroud C, Makar P, Karl T, Guenther A, Geron C, Turnipseed A, Nemitz E, Baker B, Potosnak M,
Fuentes JD. Role of canopy-scale photochemistry in modifying biogenic-atmosphere exchange
of reactive terpene species: results from the CELTIC field study. J Geophys Res Atmos. 2005;
110(D17). doi:10.1029/2005jd005775
VanReken TM, Ng NL, Flagan RC, Seinfeld JH. Cloud condensation nucleus activation properties
of biogenic secondary organic aerosol. J Geophys Res Atmos. 2005;110(D7):D07206.
Warneck P. Chemistry of the natural atmosphere. 2nd ed. San Diego: Academic; 2000.
19
599
Warneck P, Williams J. The atmospheric chemists companion. New York: Springer; 2012.
doi:10.1007/978-94-007-2275-0. 436 pp.
Went FW. Blue hazes in the atmosphere. Nature. 1960;187(4738):6413.
WHO. Air quality guidelines global update 2005. Geneva: World Health Organisation; 2005.
Further Reading
Bender J, Weigel HJ. Changes in atmospheric chemistry and crop health: a review. Agron Sustain
Dev. 2011;31(1):819. doi:10.1051/agro/2010013.
Online version available at http://www.knovel.com/web/portal/browse/display?_EXT_
KNOVEL_DISPLAY_bookid2126&VerticalID0
Penuelas J, Staudt M. BVOCs and global change. Trends Plant Sci. 2010;15(3):13344.
doi:10.1016/j.tplants.2009.12.005.
20
Contents
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Biomass Energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Biomass Conversion Technologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Bioenergy Feedstocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
The Case for Liquid Biofuels in the World Energy Market . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Ecological Considerations Associated with Cellulosic Biofuel Production . . . . . . . . . . . . . . . . . .
Management Decisions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Greenhouse Gas Emissions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Soil and Nutrient Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Impacts on Wildlife and Biodiversity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Invasive Species Potential . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Pests and Pathogens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Future Directions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
602
605
605
611
614
615
615
617
619
622
623
624
626
627
627
Abstract
K. OKeefe (*)
Division of Biology, Kansas State University, Manhattan, KS, USA
e-mail: kokeefe@k-state.edu
C.J. Springer
Department of Biology, Saint Josephs University, Philadelphia, PA, USA
e-mail: cspringe@sju.edu
J. Grennell S.C. Davis
Voinovich School of Leadership and Public Affairs, Ohio University, Athens, OH, USA
e-mail: jg509012@ohio.edu; daviss6@ohio.edu
# Springer Science+Business Media New York 2014
R.K. Monson (ed.), Ecology and the Environment, The Plant Sciences 8,
DOI 10.1007/978-1-4614-7501-9_9
601
602
K. OKeefe et al.
Introduction
Nonrenewable natural resources such as coal, petroleum, and natural gas have long
been exploited for energy consumption due to their historic relative abundance,
versatility, transportability, and low cost. However, global reserves of these raw
materials are finite and are rapidly decreasing as global demand for energy increases.
20
603
Extracting and using these energy sources can also have many negative environmental consequences. For example, fossil fuel combustion releases geologically stored
carbon and other pollutants into the atmosphere, including greenhouse gases that
cause climate change, indirectly damage ozone, contribute to acid deposition,
and cause ocean acidification (Schlesinger and Bernhardt 2013). The physical
exploitation of these fuels also damages the earths surface layers, contaminates
watersheds, and occasionally results in accidental marine contamination. Overall, the
depletion of fossil fuel reserves, increasing global demand for energy, and
the adverse environmental impacts associated with liquid and solid fossil fuel
exploitation have highlighted the need to decrease dependence on nonrenewable
fuel sources and have stimulated global interest in replacing fossil energy with
alternative, sustainable solutions for future energy consumption.
Renewable energy technologies such as solar power, wind power, geothermal
power, and bioenergy have the potential to improve energy sustainability and
reduce the environmental consequences associated with human energy consumption. Bioenergy, in particular, is a renewable energy source that is primarily derived
from plant material and is used to produce various energy products via direct
combustion or chemical processing. Bioenergy feedstocks (i.e., biomass) include
dedicated energy crops, agricultural food crops and residues, oil products, and other
organic waste materials. Depending on the raw material and conversion pathway
used, these feedstocks can produce an array of energy products ranging from liquid
biofuels (e.g., biodiesel and bioethanol) to heat and electricity. Bioenergy is widely
regarded as a viable alternative energy source because it has the potential to offer a
broad range of socioeconomic and ecological benefits. In addition to reducing
reliance on traditional fossil fuels, biomass production and biofuel processing can
create employment opportunities, particularly in rural areas, and provide energy
independence in both developing and industrialized countries. Bioenergy may also
reduce carbon emissions because bioenergy crops sequester atmospheric carbon
dioxide as they grow and because biomass combustion only releases as much
carbon dioxide into the atmosphere as plant growth has sequestered. Therefore,
bioenergy has the potential to become an economically beneficial and environmentally sustainable solution to the present energy crisis.
Although bioenergy is versatile and can provide various solutions to current
energy concerns, biomass is predominantly used in the developed world for biodiesel or bioethanol to replace petroleum transportation fuels. These fuels are of
particular interest because they do not require major modifications to current
transportation systems and can be easily mixed with fossil petroleum as fuel
additives. Presently, biofuels are produced from first-generation (i.e., conventional) sources that are also used commercially as food crops. For instance,
bioethanol is fermented from sugar sources such as corn grain (Zea mays L.) or
sugar cane (Saccharum officinarum L.), while biodiesel is processed from oil crops
such as soybean (Glycine max, L.) and rapeseed (Brassica napus L.). The technologies used to produce first-generation biofuels are currently well established, and
although biofuel additives/substitutes are not yet major energy sources in the
transportation sector, their production is now commercial. Biofuels do have the
604
K. OKeefe et al.
20
605
disadvantages of advanced cellulosic biofuels as an alternative fuel source, particularly with respect to production efficacy and environmental impacts.
Biomass Energy
Plants use solar radiation, carbon dioxide, water, and mineral nutrients to convert solar
energy into chemical energy through the process of photosynthesis. In plants, this
energy is stored primarily in the form of soluble and non-soluble carbohydrates that
drive metabolic activity and form tissue structures, respectively. When carbohydrate
bonds are broken via industrial conversion technologies, the energy released can be
captured and used as a source of fuel for human consumption. If combusted, the
energy takes the form of heat and can be used to produce electricity. If chemically
converted to liquid hydrocarbons, the energy remains in a chemical form that can fuel
combustion engines. In these processes, the plant tissue is referred to as biomass.
Biomass has been a steady and reliable source of heat throughout human history and
remains so in some developing nations where biomass-generated heat comprises up to
90 % of energy consumption. In developed nations, biomass energy delivers a
significantly lower (~34 %) proportion of the total energy consumed (Demirbas
2009). However, there has been a concerted effort in recent years to increase the
contribution of biomass energy to national energy budgets, particularly from advanced
cellulosic sources. Currently, the United States leads the world in bioenergy production mainly due to the use of ethanol in blended fuels (Fig. 1). In the next few sections,
the technology necessary for the conversion of biomass to liquid cellulosic fuels, the
fuel products generated from these processes, and the most common plant species used
for liquid cellulosic fuel production will be summarized, as will the potential for
biomass energy to contribute to the global energy supply in the future.
606
K. OKeefe et al.
20
607
Fig. 2 Primary approaches used to convert lignocellulosic biomass into bioenergy via (a)
biochemical and (b) thermochemical conversion processes. Also indicated are the most common
fuel derivatives from each process
608
K. OKeefe et al.
Four basic steps are required for the biochemical production of liquid cellulosic
ethanol: size reduction, pretreatment, hydrolysis, and fermentation. The first, size
reduction, is the mechanical reduction of biomass to a smaller size; this facilitates
the access of reagents used in the fuel generation process to the array of carbohydrates found in cellulosic biomass. Once the initial feedstock is mechanically
processed, the complex lignocellulosic molecules must be broken down into fermentable material via hydrolysis and, depending on the type of hydrolysis used,
pretreatment prior to hydrolysis.
Defined generally, hydrolysis is the process of splitting a molecule into smaller
fragments with the addition of water. In the biochemical conversion of biomass to
liquid biofuel, hydrolysis is used to cleave (depolymerize) complex lignocellulosic polymers (cellulose and hemicellulose) into simple constituent sugar monomers (glucose, pentose, hexose, and xylose) that can subsequently be fermented
into ethanol fuel. Hydrolysis of cellulosic biomass can occur chemically or
enzymatically. Chemical, or acid hydrolysis, uses an acid (typically sulfuric
acid or hydrochloric acid) in the presence of water to break down cellulose into
glucose and hemicellulose into pentoses and hexoses. Xyloses can also be produced from hardwood and crop residue feedstocks. Lignin, however, is very
resistant to degradation and therefore remains as a by-product in these reactions.
Additional toxic by-products, such as hydrolyzates, may also form during this
process.
Although acid hydrolysis does not require pretreatment beyond the mechanical
processing of the raw biomass feedstock, this process is costly and energetically
expensive because the sugar products must be conditioned before fermentation to
remove toxic by-products and also because the acid used for hydrolysis must be
recovered. Therefore, enzymatic hydrolysis is the more common process used to
depolymerize cellulosic biomass into simple sugars.
If carried out enzymatically (enzymatic hydrolysis), the biomass feedstock must
first undergo pretreatment. Pretreatment is the physical, chemical, or enzymatic
degradation of biomass that is used to increase enzyme access to cellulose and other
polysaccharide components of the biomass feedstock. This typically involves the
breakdown of the biomass with the same chemicals used during acid hydrolysis
(sulfuric acid or hydrochloric acid), which results in the partial hydrolysis of the
biomass. After the biomass has been pretreated, the remaining material that has not
been degraded by the acid can then undergo hydrolysis catalyzed by a mixture of
enzymatic compounds. This process requires the use of cellulases, a class of
enzymes derived from bacterial or fungal sources.
Following acid or enzymatic hydrolysis, the simple sugars, mostly xylose
(derived from woody species) and glucose (derived from non-woody species), are
converted to liquid bioethanol fuels through fermentation. Fermentation reactions
can occur under aerobic or anaerobic conditions and are driven by many different
kinds of microorganisms in nature. In the processing of liquid biofuel from cellulosic biomass, microorganisms use sugar monomers to produce ethanol. Once the
enzymatic fermentation steps are complete, distillation and dehydration processes
20
609
610
K. OKeefe et al.
of gas, solid carbon-based char, and liquid bio-oil fuel. The ratios of these
pyrolysis products largely depend on a number of factors including the reaction
temperature, pressure, rate of heating, the length of the reaction time, and the
biomass feedstocks utilized at the onset. The process of pyrolysis begins by
heating the biomass feedstock to a high temperature (ranging from 200 C
to >1,000 C, depending on the method). At this point volatile organic compounds, or VOCs, form and leave behind a carbonaceous char, a process known
as primary pyrolysis. The release of VOCs results in a transfer of heat from the hot
VOCs to the feedstock that has yet to be pyrolysed. As the VOCs cool, they form
tar. Finally, autocatalytic secondary pyrolysis occurs while primary pyrolysis
occurs simultaneously, leading to the production of liquid biofuel. Currently,
three types of pyrolysis reactions exist to produce char, tar, and liquid cellulosic
fuels. The first is known as slow pyrolysis. The slow rate of heating in slow
pyrolysis produces a higher ratio of char than liquid and gas products. The second,
fast pyrolysis, involves fast heating rates and results in a much higher ratio of liquid
cellulosic fuels. Finally, flash pyrolysis is a more efficient mechanism similar to
fast pyrolysis except that very high reaction temperatures and very high heating
rates of the reactions are used. Due to the extremely fast heating in flash pyrolysis,
the conversion of biomass feedstocks to fuel is much more efficient and leads to a
fuel product that does not require further refinement after the initial pyrolytic
reactions have occurred.
An additional thermochemical conversion process of biomass feedstocks
to liquid cellulosic fuels is known as liquefaction. Liquefaction is the process
of heating biomass feedstocks to low temperatures under high pressure with
the addition of a catalyst, solvent, and/or reducing gas such as hydrogen to
produce a highly viscous insoluble oil that can be used for a variety of purposes.
At this time, there is low interest in liquefaction as a viable thermoconversion
process because of the complexity and expense associated with building reactors
when compared to other thermoconversion processes like gasification and
pyrolysis.
Finally, it should be noted that the physical properties of cellulosic plant material
can often complicate thermochemical conversion processes. For instance, the high
water and oxygen content of the plant material can produce large quantities of
smoke during combustion, while the fibrous nature of its lignocellulosic cell walls
can make the biomass physically difficult to process. Recent advancements therefore recommend pretreating the biomass to increase the quality of the biomass and
to reduce undesirable side effects associated with fuel production. Torrefaction is a
pretreatment method that is similar to pyrolysis but occurs at much lower temperatures (200300 C). This process removes oxygen from the plant tissue and
decreases the volume of the tissue by as much as 6269 %. In doing so, the energy
content of the biomass is maintained because the material dries and partially
de-volatilizes. This reduction in biomass and concomitant energy preservation
can increase the energetic density of the material by approximately 2030 %,
which not only makes the material easier to process but also aids in transportation
(Bhagwan Goyal et al. 2009).
20
611
Table 1 Major environmental impacts and considerations for first-generation and advanced
cellulosic bioenergy feedstocks
Feedstock type Example feedstocks
First
Sugar crops
generation
Zea mays (corn)
Saccharum officinarum (sugar
cane)
Advanced
Bioenergy Feedstocks
The plant species that can be grown as cellulosic bioenergy crops are even more
diverse than the processing technologies used to produce biofuels (Table 1). Many
crops and wild plant species are currently being used as bioenergy feedstocks, are in
development to be used to produce biofuel, or are excellent candidates for biofuel
production in the future. Examples of such feedstock plants are agricultural wastes,
trees wastes and residues, food crops, and perennial grasses. Generally, these plant
species are classified into two categories: food crops and bioenergy crops. Plant
species associated with each of these groups present unique challenges in the
production of suitable biomass for liquid cellulosic fuel manufacturing and also
have varying environmental concerns linked to their growth. A number of characteristics need to be considered when deciding which species to use as a feedstock,
including mineral content, moisture content, nutrient and water requirements, dry
matter production per unit land area, and the chemical composition of the tissue,
especially lignin, hemicellulose, and cellulose content. Furthermore, the geographical distribution of the plant species, the effects of the species on the environment
612
K. OKeefe et al.
20
613
Fig. 3 Panicum virgatum L. (a) and Miscanthus gigantea (b), two species of grass with a high
potential to become important feedstocks for second-generation biofuels (Photo credit: Kimberly
OKeefe (a) and Sarah Davis (b))
below the yields generated by corn. However, its extensive root system,
low-nutrient and water requirements, and perennial lifecycle make switchgrass an
attractive option as a biomass feedstock for conversion to bioethanol. In addition,
the loss of biomass during harvest is low for switchgrass. Also increasing the
attractiveness of switchgrass as a biomass feedstock are the genetic tools, such as
linkage maps, that have been developed in recent years for use in breeding programs, an effort that results from the intense focus on switchgrass by the United
States Department of Energy (Perlack et al. 2005).
Miscanthus x giganteus is another intensely researched option for biofuel production. Miscanthus x giganteus is also a rhizomatous, C4 perennial grass that is a sterile
hybrid of Miscanthus saccharifloris and Miscanthus sinensis, both native to Asia.
Currently a single sterile hybrid, Miscanthus x giganteus Greef et Deuter, is the
cultivar primarily studied for biofuel production. Like switchgrass, Miscanthus x
giganteus produces a large aboveground component and stores much of its nutrients
belowground in rhizomes prior to harvesting, thus reducing the nutrient requirements
for the species. In fact, a number of studies have found Miscanthus x giganteus
productivity to largely be unresponsive to nitrogen additions. Also like switchgrass,
Miscanthus x giganteus has been successfully grown in a number of locations globally
and therefore represents a single product for conversion technologies to use as a
biomass feedstock. Miscanthus x giganteus is generally more productive than switchgrass, yielding 2530 t/ha annually, but can also have greater loss of biomass at harvest
614
K. OKeefe et al.
than switchgrass. One major unknown for Miscanthus x giganteus productivity that
requires additional research is the effect of environmental conditions on the productivity of the species. Other challenges presented by Miscanthus x giganteus as a
biomass feedstock are related to the self-incompatible nature of the species. This
self-incompatibility has made genetic research on the species challenging and has
hampered genetic improvements to date (but this also what eliminates the invasion
risk of Miscanthus x giganteus). Recently, a low-density genetic map has been
produced for Miscanthus x giganteus and has resulted in some genetic studies aimed
at the heritability of select agronomic characteristics.
Other advanced dedicated bioenergy feedstocks that are currently under development are woody plant species such as Populus spp. (poplar) and Salix spp.
(willow), also known as short-rotation woody crops (SRWCs). Of the woody
plant species being considered for use as a biofuel feedstock, hybrid poplar received
the most attention because it has the potential to yield biomass for conversion at a
high rate. For example, yields of poplar are estimated between 12.4 t/ha on
nonirrigated and unfertilized land to 22.5 t/ha on land that has been irrigated and
fertilized. Poplar is also attractive because a number of genomic and genetic tools
such as a fully sequenced genome are available for use by the existing research
community. However, major setbacks associated with the use of poplar as a biofuel
feedstock is the long generation time of the plant as well as the long-term sustainability of yields under low nitrogen inputs to the soil. Engineering the species for
increased yields and decreased nitrogen requirements will therefore be important
steps in developing poplar as a sustainable alternative energy solution. These
improvements have the potential to also enhance production for the timber and
paper industries as well (Field et al. 2008).
Finally, a readily available cellulosic feedstock can be gathered from agricultural waste products. Agricultural wastes include corn stover (the leaves and stalks
of the corn crop), sugarcane residues, and rice hulls, as well as a number of other
agricultural residues. Because of the inefficiencies that exist in current conventional
agriculture production, these waste products remain among the most important
feedstocks for biofuel production. In addition, the use of agricultural wastes is
generally thought to be a better option for biofuel production than food crops such
as corn because these waste products are the by-products of existing agricultural
activity. This reduces the energy requirement needed for production as well as the
use of pesticides, fertilizers, and water. In many cases, if agricultural wastes are not
used for biofuel production, they are either burned or disposed in a landfill, both of
which can have a higher environmental impact than the production of biofuel.
Additionally, forest residues from logging, as well as municipal wastes, also have
the potential to be used as cellulosic bioenergy feedstocks.
20
615
increase, the case for liquid fuels produced from sustainable sources becomes
stronger. Many factors contribute to this reality. First, because of the wide geographic ranges of the biomass feedstocks described above, biofuels present an
opportunity for increased domestic energy production. This new area for domestic
energy production also creates a unique opportunity for economies to further
develop agricultural industries and buildup rural communities. In addition to
providing energy security, biofuels have the ability to decrease greenhouse gas
emissions by reducing the use of fossil fuels while simultaneously increasing the
potential for long-term sequestration of atmospheric carbon dioxide in plant tissues
and soils.
Another attractive feature of liquid biofuels is the amount of energy, especially
in the transportation sector, that they can generate currently and in the future. The
earth receives annually ~3.8 106 exajoules (1 EJ 1018 J) of solar energy. This
is such an extraordinary amount of energy that it meets the total global annual
energy demand (450 EJ year1) in 1 h of daylight. Globally, plants fix many times
more (~2,900 EJ year1) than the total annual global energy demand by converting
this solar energy into standing biomass through photosynthesis, a term known as net
primary productivity (NPP). Unfortunately, not all of this energy is available for use
as bioenergy feedstocks on a sustainable basis. Of this total, a number of estimates
have placed the total energy potential from sustainable biomass feedstocks at
100300 EJ year1. Currently, only 4055 EJ year1 of energy is produced from
biomass globally. Because of the significant potential for much higher proportions,
a number of developed nations have committed to significantly increasing the
amount of bioenergy used by the year 2050. The Intergovernmental Panel on
Climate Change estimates that by the year 2050 global energy demand will increase
to ~560 EJ year1. Currently, the energy generated from biomass has the potential
to meet 32 % (180 EJ year1) of the future global energy demand laid out by the
IPCC. This proportion is projected to increase to 46 % (325 EJ year1) by 2100.
While the use of energy derived from biomass has the potential to change the global
energy portfolio, there are a number of pressing environmental and sustainability
considerations that must be accounted for now and in the future (Field et al. 2008).
616
K. OKeefe et al.
cellulosic bioenergy production will impact the environment will depend on the
crops used, the land-use changes associated with them, and how these crops are
managed. If managed poorly, their production may not provide any ecological
benefits over first-generation bioenergy crops and may even adversely impact the
environment. When evaluating the ecological consequences associated with cellulosic bioenergy production, the following factors should be considered.
Land-Use Change
Bioenergy production on a scale large enough to meet current energy needs will
require substantial areas of land provided via some form of land-use change (LUC)
(Davis et al. 2011a, b). Land-use changes can directly alter existing land (e.g.,
agricultural land used previously for the production of other crops, natural ecosystems converted to agricultural land, or marginal land that is degraded from extensive agricultural practice and is unsuitable for further food crop production).
Bioenergy cultivation can also indirectly induce land-use changes (i.e., indirect
land-use change, iLUC), when uncultivated land is altered to produce a crop that
was displaced by bioenergy feedstock cultivation on current agricultural land. For
instance, in a hypothetical scenario where biofuel production replaces wheat production on a farm, the farmer may convert a native ecosystem in another area to
compensate for the lost wheat production in the original location.
When addressing iLUC, it is important to note that tracking and predicting the
many variables associated with iLUC is extremely difficult and associated with
large uncertainty. Some agencies, such as the United States Environmental Protection Agency (EPA), have tried to evaluate effects of iLUC using models, but the
models produce results with large variance because accounting for all associated
variables can be difficult (Davis et al. 2011a). Although estimating the effects of
iLUC is difficult, most models indicate that there are unintended consequences of
biofuel policies for land use that should be addressed. Because iLUC has the
potential to be counterproductive to mitigating climate change (through the development of uncultivated land), awareness of this potential consequence is key for
policy makers to place regionally appropriate constraints on biofuel development.
Evidence is mounting that integrated land management policies might reduce
unintended consequences for LUC and iLUC (Davis et al. 2011a). As discussed
previously, many opportunities exist to coproduce biofuels and other resources,
reducing the need for additional land. Wastes from many industries could serve as
biofuel feedstocks, including residues from the timbering, paper, wood products,
building, and agricultural industries. However, opportunities for integrated land
management are often regionally specific and cannot be generalized globally for
similar industries.
Crop Management
Various management decisions associated with annual first-generation biofuel crop
production, such as planting, harvesting, and tilling methods, can also determine the
environmental impact of a bioenergy crop (Howarth and Bringezu 2009). For
example, bioenergy crops can be planted as monocultures (single-species crop
20
617
stands) or in mixed assemblages (multiple species are planted in crop stands). They
can also be planted in crop rotations, where different crops are planted and
harvested on a rotational schedule (e.g., crop A is grown and harvested for
5 years and then crop B is grown and harvested for 5 years). The timing and pattern
of crop harvest can also vary. Crops can be harvested annually, more than once per
growing season, or less than once per growing season. The entire crop stand can be
harvested at once, or alternatively, only sections of the stand can be harvested at
once (i.e., strip harvesting). Crops can be harvested during different seasons.
Finally, agricultural soil can be tilled, or mechanically disturbed to facilitate crop
planting, in a variety of ways. Intensive tillage methods leave few crop residues,
whereas less intensive, or reduced, tillage methods leave greater amounts of crop
residues. No-till strategies do not till agricultural soil prior to planting, which leaves
crop residues undisturbed. Strip-till methods only disturb the soil directly where the
crop is planted, leaving strips of untilled soil between. Rotational tillage only tills
the soil at particular intervals (i.e., every other year). Variation in any of these
factors will ultimately influence the degree to which cellulosic bioenergy crops
affect greenhouse gas emissions, soil properties, soil and water quality, and
wildlife.
618
K. OKeefe et al.
reduces soil organic carbon (SOC) and releases CO2 into the atmosphere (see
section Soil Organic Carbon). The magnitude of CO2 release, however, depends
on the type of land-use change used to cultivate the bioenergy crop. Land-use
changes release more CO2 if highly productive land, such as a forest ecosystem, is
converted to a bioenergy crop field than if the conversion occurs on agricultural land
or marginal land (a low productivity ecosystem) (Davis et al. 2011b). The bioenergy
feedstock chosen for cultivation can also determine the impact of bioenergy production on greenhouse gas emissions over time. Perennial grasses, for instance, can
grow for many years without the need to till and replant, resulting in greater
accumulation of SOC relative to an annual cropping system (Blanco-Canqui 2010).
Increased SOC sequestration in dedicated herbaceous bioenergy crops relative to
first-generation bioenergy crops can therefore create a net greenhouse gas sink if this
perennial system replaces annual corn agriculture. A popular metric that is used to
determine if land-use change results in positive or negative consequences for
ecosystem services is the payback time needed to neutralize the carbon debt incurred
through soil disturbance and the removal of vegetation from the landscape (Davis
et al. 2011a). The payback time is dependent on the original condition of the land
(e.g., soil, aboveground plant community, management history), climate, and the
rate at which the biofuel agroecosystem sequesters carbon.
Bioenergy production can also release N2O emissions if substantial fertilizer
inputs are used to grow the crop. Fertilizers add nitrogen to the soil in the form of
ammonium (NH4+), which can increase rates of microbial nitrification and denitrification in the soil and subsequently produce gaseous nitric oxide (NO) and N2O as
by-products (Schlesinger and Bernhardt 2013). N2O is highly inert and has a long
residence time in the atmosphere, so it has great potential to warm the atmosphere
over time (about 300 greater than atmospheric CO2). N2O also breaks down into
NO when exposed to ultraviolet radiation in the stratosphere, which can promote
stratospheric ozone destruction and subsequently increase the amount of harmful
solar radiation that reaches the surface of the planet. Therefore, N2O production
associated with agricultural activities can have wide-ranging consequences for
atmospheric chemistry and climate. Cellulosic feedstocks are generally less likely
to produce N2O emissions than first-generation bioenergy crops because dedicated
herbaceous bioenergy crops are often characterized by high nutrient-use efficiency
and can sometimes be grown without the addition of nitrogen fertilizer (see section
Nutrient and Chemical Inputs) (Williams et al. 2009). Low nitrogen inputs reduce
rates of nitrification and denitrification in the soil, which can ultimately reduce N2O
emissions. Dedicated herbaceous bioenergy crops can also be grown under drier
conditions than traditional row crop monocultures, which may reduce rates of
denitrification and reduce N2O emissions compared to first-generation bioenergy
crops. However, soil disturbance associated with land conversion can also accelerate nitrogen cycling processes, which may increase N2O emissions associated with
the establishment of a dedicated herbaceous bioenergy crop despite their
low-nutrient requirements. Therefore, bioenergy feedstock, land-use, and crop
management must all be considered when assessing the impact of biofuel production on terrestrial N2O emissions.
20
619
620
K. OKeefe et al.
species only replanted every 1520 years, which reduces the degree of management
by heavy equipment and thus reduces the risk of soil compaction. Reduced erosion
can benefit agricultural and natural systems by maintaining soil structure, retaining
soil organic matter and nutrients, and reducing the transport of undesirable nutrients
and/or contaminants to other natural systems (e.g., nutrient deposition and in
aquatic systems).
The degree to which cellulosic bioenergy crops reduce soil erosion depends on
the crop used and how the crop is managed (Williams et al. 2009). Runoff and
erosion are generally reduced by perennial C4 grasses and short-rotation woody
crops. Conversely, harvesting annual crop residues such as corn stover may actually
exacerbate the rate of surface runoff and soil erosion in an agricultural system
because residue removal exposes soil to wind and rainfall and the heavy equipment
used to remove the residues can compact the soil. Harvesting the crop during the
winter or when the soil is dry can however reduce soil compaction. Minimum or
no-till farming, as well as contour plowing (plowing along the landscapes elevation contour to form furrows that capture water), can also reduce surface runoff and
erosion. Cellulosic bioenergy crops that are managed more intensely (i.e., are
harvested multiple times throughout the year or are extensively tilled) can also
counteract the benefits of perennial grasses on soil structure. The degree to which
soil erosion is reduced by cellulosic bioenergy crops can depend on the type of soil
in which the crop is growing, as well as on the length of time following establishment. Perennial C4 grasses, for instance, may not reduce erosion in the first year
they are planted. In fact, these crops may not improve soil structure or soil hydraulic
properties for many years after they are established (Howarth and Bringezu 2009).
Therefore, cellulosic bioenergy crops do have the potential to reduce surface runoff
and soil erosion, although this depends on crop management and may take decades.
20
621
The low chemical inputs required for cellulosic bioenergy crop cultivation can
provide several benefits to the environment. First, low fertilizer inputs can greatly
reduce energy consumption because the production of industrial nitrogen fertilizer
(i.e., industrial nitrogen fixation via the Haber-Bosch process) is an energetically
expensive process. Second, low fertilizer, herbicide, and pesticide inputs can
improve soil quality and reduce the amount of chemicals that are present in
surface runoff, thus reducing rates of nitrification and denitrification (see section
Greenhouse Gas Emissions) and harmful ecological processes such as nitrogen
leaching and eutrophication (see section Water Quality). Low fertilizer inputs, for
example, can reduce nitrogen leaching in the soil by reducing rates of nitrification.
Nitrification is the two-step process by which aerobic chemoautotrophs oxidize
ammonium (NH4+) to nitrite (NO2) and then nitrate (NO3), a highly soluble form
of nitrogen (Schlesinger and Bernhardt 2013). Increasing NH4+ inputs to a
system via fertilization increases rates of nitrification and ultimately increases the
concentration of soluble NO3 that can leach through the soil and contaminate
groundwater. Thus, bioenergy crops that require low nitrogen inputs will
reduce NO3 leaching associated with agricultural practices. Proper management
regimes have the potential to enhance these environmental benefits. For instance,
more nutrients can be retained in the soil by harvesting biomass after plant
senescence, when nutrients have been translocated belowground to roots. Planting
crop stands in mixed assemblages with nitrogen-fixing plant species interspersed
among the biomass crop may also reduce the need for additional nitrogen input.
622
K. OKeefe et al.
organic carbon back to the soil in the form of aboveground residues and root
dieback. The extensive root systems produced by these crops also grow deep into
the soil profile, which transfers organic carbon to deep soil layers where SOC
decomposition rates are low. Thus, carbon inputs to the soil may be larger than
carbon outputs, increasing SOC over time. Increasing SOC is highly beneficial in an
agricultural system; higher SOC levels can improve soil structure, buffer soil
acidity, increase crop quality and productivity, increase the abundance of soil
microorganisms, reduce runoff, and improve water quality (U.S. Congress Office
of Technology Assessment 1993). However, the amount of SOC that bioenergy
crops can add to a system depends on a variety of factors, including soil type,
climate, and land management. The amount of carbon that feedstocks can sequester
and add to the soil also depends on the amount of carbon already present in the soil
because soil can eventually become saturated with carbon. Although absolute limits
are debated, greater amounts of carbon can be added to degraded soil that is carbondepleted than highly productive soil that is closer to its carbon saturation point
(Blanco-Canqui 2010). These crops therefore have greater potential to improve
marginal lands compared to more productive lands. The amount of organic carbon
that is added to the soil by bioenergy crops depends on the crop used and the way
the crop is managed. Perennial C4 grasses and short-rotation woody crops tend to
increase SOC, but removing annual crop or forest residues actually decreases SOC
by directly removing organic material from the soil and by exposing the soil to
higher air temperatures that increase rates of organic material decomposition
(Lemus and Lal 2005; Williams et al. 2009). Greater amounts of SOC are also
retained in the soil when crops are harvested less frequently and minimum or no-till
farming regimens are used.
Water
Water Requirements
Agricultural crops, including food crops and first-generation bioenergy crops, can
be characterized by low water-use efficiency (they assimilate less carbon per unit
water transpired) and are sometimes irrigated with water collected from lakes,
rivers, and groundwater to produce higher yields. This can have negative socioeconomic and environmental consequences because irrigation aggravates water shortages and reduces surface water flow necessary for wetland ecosystems and aquatic
biota. Many dedicated herbaceous bioenergy crops can produce high yields without
irrigation because these perennial grasses utilize the C4 photosynthetic pathway and
use water more efficiently than plants that utilize the C3 photosynthetic pathway
(Carroll and Somerville 2009; Williams et al. 2009). In addition, many perennial
bioenergy feedstocks have extensive deep root systems that aid in retaining water in
the soil more than the small root systems associated with annual row crops, further
reducing the need for irrigation (Howarth and Bringezu 2009). Because they do not
require as much irrigation, cellulosic bioenergy feedstocks compete less with
food crops for water and are also less likely to impact aquatic systems than
20
623
Water Quality
Cellulosic bioenergy feedstocks can also improve water quality relative to annual
row crops. These crops do not require substantial chemical inputs, and their
extensive root systems, as well as their ample SOC inputs, reduce surface runoff
and soil erosion. This can decrease chemical contamination of aquatic habitats and
can subsequently reduce nitrogen leaching (see section Nutrient and Chemical
Inputs) and aquatic eutrophication (i.e., aquatic ecosystem responses to nutrient
additions). Thus, these feedstocks are less associated with negative aquatic processes such as phytoplankton or algal blooms and hypoxic conditions (oxygen
depletion) than annual crops (Blanco-Canqui 2010).
624
K. OKeefe et al.
heterogeneity and create a more diverse environmental mosaic that can support
more species in a given area. Harvesting strategies may also affect the degree to
which bioenergy crops impact biodiversity (Fargione et al. 2009). Frequent harvests
(>1 harvest per year) may favor species that prefer a short-grass habitat, while
infrequent harvests may favor species that prefer tallgrasses. Rotational or strip
harvesting can improve environmental heterogeneity and support the coexistence of
multiple species that prefer different habitats. Crop harvests can also interfere with
avian breeding seasons, so harvesting in the autumn or winter, after the breeding
season of many bird species has ended, may benefit a variety of bird species
(Dauber et al. 2010). However, autumn or winter harvests can reduce ground
cover and consequently increase winter mortality for many ground-dwelling birds
and mammals. Crop management strategies can therefore have wide-ranging
impacts on many wildlife species, and these consequences must be carefully
considered when making land management decisions to cultivate cellulosic
bioenergy crops.
Other cellulosic bioenergy crops may also impact wildlife and biodiversity. For
example, short-rotation woody crops can provide habitat for birds and small
mammals, although these habitats are often less suitable than natural forests
because crop stands are less complex than naturally occurring forest ecosystems
(Dauber et al. 2010). Woody crops may also reduce habitat fragmentation if
planted as a corridor to connect separated forest patches. Reduced habitat fragmentation can facilitate the movement of individuals and populations between
habitats and is ultimately associated with high biodiversity. Finally, annual crop
residues, as well as forest residues, tend to have fewer impacts on biodiversity than
short-rotation woody crops or perennial grasses because their collection is not
associated with land-use changes that reduce viable habitat or environmental
heterogeneity. Residue removal, however, does reduce ground cover for wildlife
and also decreases soil nutrients, which may impact the biodiversity of grounddwelling animals or soil microorganisms. Therefore, the type of bioenergy feedstock used, as well as the strategy used to plant and maintain the crop, can strongly
influence the degree to which cellulosic bioenergy cultivation impacts wildlife and
biodiversity.
20
625
Interestingly, these characteristics are also those associated with many cellulosic
bioenergy crops. Dedicated herbaceous bioenergy crops, for example, are perennial,
have rapid growth rates, produce high yields, utilize the C4 photosynthetic pathway,
have high resource-use efficiency, and propagate both vegetatively and by producing a seed crop. These species are also broadly adapted across a wide geographic
range and are tolerant of various environmental conditions. For instance, crops such
as switchgrass and Miscanthus x giganteus are tolerant of drought, as well as
flooding, and can grow on low-nutrient, degraded soils. These traits promote efficient seedling establishment and quick production of high yields with relatively little
water and nutrient inputs. However, these traits may also promote the undesirable
invasion of bioenergy crops into nonagricultural areas, particularly if the crop is
cultivated outside of its native range or if the crop is genetically modified to enhance
qualities that concomitantly increase invasiveness (Raghu et al. 2006; Williams
et al. 2009). Unintentional introduction can occur locally or on larger scales as a
result of direct spread from the agricultural land source or by propagule release
during harvesting and processing (Fargione et al. 2009). Biomass feedstocks are
typically harvested following plant senescence, when seeds have been produced and
are still attached to the plant, which can result in seed rain onto roadsides during
transportation to biofuel production facilities. These seeds may also contaminate the
equipment used to plant or harvest the crop, which may subsequently taint other
agricultural crops if the equipment is not properly sterilized.
A non-native bioenergy crop may survive and form persistent populations
because it will likely experience a decrease in pressure from specialist enemies
(i.e., specialist pathogens and herbivores) when introduced to a new region (Hierro
et al. 2005). The non-native species is not typically susceptible to the specialist
enemies of the native species in its new range (assuming that these specialist
enemies do not switch host preference to the invader) and should therefore experience a decrease in regulation by enemies relative to the native species in the new
region. The risk of invasion, however, may decline if native crops or sterile
cultivars (such as Miscanthus x giganteus) are cultivated, although other traits
associated with these species may promote their invasiveness despite their lack of
a viable seed crop. Invasiveness is not typically associated with other cellulosic
bioenergy sources such as annual crop residues and short-rotation woody crops.
626
K. OKeefe et al.
risk of invasion. Growing multiple genotypes of a single species may also increase
landscape heterogeneity and reduce invasions by other species. Similarly, other
species may become invasive in bioenergy agricultural lands if substantial amounts
of water and/or nutrients are added to the crop, which may create more available
resource niches that can potentially be utilized. Most cellulosic bioenergy crops,
however, do not require substantial water or nutrient inputs, so this risk may
actually be lower compared to traditional row crops.
20
627
Future Directions
Biofuels produced from cellulosic sources have the potential to reduce the need for
fossil fuel energy in the future. As mentioned throughout the previous sections,
advanced cellulosic crops generally possess many ecological advantages over firstgeneration feedstocks. Cellulosic bioenergy crops typically have extensive rooting
systems and produce high yields without requiring large water or nutrient inputs.
These characteristics can increase SOC and reduce rates of greenhouse gas emissions, runoff, and eutrophication. Additionally, perennial cellulosic crops can be
cultivated on land considered marginal for agricultural production and improve
wildlife habitat quality. However, cellulosic feedstocks are also difficult to process
into liquid fuel, and depending on how they are managed, their impact on natural
ecosystems may not always be positive. Therefore, the development of cellulosic
biofuels for widespread future production requires continued research in areas of
feedstock propagation and conversion technologies. Specifically, future work in the
development of biofuels from cellulosic sources should aim to improve the conversion of cellulosic biomass to liquid fuel. Genetically engineering bioenergy crop
species to make lignocellulosic material easier to hydrolyze, either by reducing or
modifying lignin content, may increase the cost efficiency of liquid biofuel production. Work in the future should also focus on the development of new enzymes
that are better able to break down lignocellulosic biomass. Ultimately, these
developments will require more research to better understand plant cell wall
chemistry. A better understanding of the environmental impacts associated with
bioenergy feedstock production is also needed. Bioenergy crops can have various
impacts on the environment, depending on the crop used and how the crop is
managed, so predicting how bioenergy production will impact various ecosystems
in the future can be difficult. This will be especially important in the face of global
climate change, as different crops will likely respond differently to changes in
atmospheric chemistry and climate. Therefore, if the full benefits of cellulosic
bioenergy production are to be realized, a dedication must be made to the production and management of bioenergy feedstocks that not only have few adverse
impacts on the environment but that are also more efficient in generating liquid
fuels from lignocellulosic material.
References
Bhagwan Goyal H, Saxena RC, Seal D. Thermochemical conversion of biomass to liquid and
gaseous fuels. In: Pandey A, editor. Handbook of plant-based biofuels. Boca Raton: CRC
Press; 2009. p. 2943.
Blanco-Canqui H. Energy crops and their implications on soil and environment. Agronomy
J. 2010;102:40319.
Carroll A, Somerville C. Cellulosic biofuels. Annu Rev Plant Biol. 2009;60:16582.
Dauber J, Jones MB, Stout JC. The impact of biomass crop cultivation on temperate biodiversity.
Glob Change Biol Bioenerg. 2010;2:289309.
Davis SC, House JI, Diaz-Chavez RA, Molnar A, Valin H, DeLucia EH. How can land-use
modeling tools inform bioenergy policies? J R Soc Interf Focus. 2011a;1:21223.
628
K. OKeefe et al.
Davis SC, Parton WJ, Del Grosso SJ, Keough C, Marx E, Adler P, DeLucia EH. Impacts of
second-generation biofuel agriculture on greenhouse gas emissions in the corn-growing
regions of the US. Front Ecol Environ. 2011b;10:6974.
Demirbas MF. World biofuel scenario. In: Pandey A, editor. Handbook of plant-based biofuels.
Boca Raton: CRC Press; 2009. p. 1328.
Fargione JE, Cooper TR, Flaspohler DJ, Hill J, Lehman C, McCoy T, Nelson EJ, Oberhauser KS,
Tilman D. Bioenergy and wildlife: threats and opportunities for grassland conservation.
BioScience. 2009;59:76777.
Gonzalez-Hernandez JL, Sarath G, Stein JM, Owens V, Gedye K, Boe A. A multiple species
approach to biomass production from native herbaceous perennial feedstocks. In Vitro Cell
Dev Biol Plant. 2009;45:26781.
Hierro JL, Maron JL, Callaway RM. A biogeographical approach to plant invasions: the importance of studying exotics in their introduced and native range. J Ecol. 2005;93:515.
Howarth RW, Bringezu S. Biofuels: environmental consequences and interactions with changing
land use. Proceedings of the Scientific Committee on Problems of the Environment (SCOPE)
International Biofuels Project Rapid Assessment; 2008 Sept 2225; Gummersbach, Germany.
Ithaca/New York: Cornell University; 2009.
Lemus R, Lal R. Bioenergy crops and carbon sequestration. Crit Rev Plant Sci. 2005;24:121.
Perlack RD, Wright LL, Turhollow AF, Graham RL, U.S. Department of Agriculture and
U.S. Department of Energy. Biomass as feedstock for a bioenergy and bioproducts industry:
the technical feasibility of a billion-ton annual supply. GO-102005-2135. Washington, DC:
Government Printing Office; 2005.
Raghu S, Anderson RC, Daehler CC, Davis AS, Wiedenmann RN, Simberloff D, Mack
RN. Adding biofuels to the invasive species fire? Science. 2006;313:1742.
Saxena RC, Adhikari DD, Goyal HB. Biomass-based energy fuel through biochemical routes: a
review. Renew Sustain Energy Rev. 2009;13:16778.
Schlesinger WH, Bernhardt ES. Biogeochemistry: an analysis of global change. San Diego:
Academic; 2013.
Schrotenboer AC, Allen MS, Malmstrom CM. Modification of native grasses for biofuel production may increase virus susceptibility. GCB Bioenerg. 2011;3:36074.
Taiz L, Zeiger E. Plant physiology. 5th ed. New York: Sinauer; 2010.
U.S. Congress Office of Technology Assessment. Potential environmental impacts of bioenergy
crop production-background paper, OTA-BP-E-118. Washington, DC: U.S. Government Printing Office; 1993.
Williams PRD, Inman D, Aden A, Heath GA. Environmental and sustainability factors associated
with next-generation biofuels in the U.S.: what do we really know? Environ Sci Technol.
2009;43:476375.
Further Reading
Buckeridge MS, Goldman GH. Routes to cellulosic ethanol. New York: Springer; 2011.
Burkheisser EV. Biological barriers to cellulosic ethanol. Hauppauge: Noval Science; 2011.
Canfield D, Glazer AN, Falkowski PG. The evolution and future of Earths nitrogen cycle.
Science. 2010;330:1926.
Cheng J. Biomass to renewable energy processes. Boca Raton: CRC Press; 2009.
Falkowski P, Scholes RJ, Boyle E, Canadell J, Canfield D, Elser J, Gruber N, Hibbard K,
Hogberg P, Linder S, Mackenzie FT, Moore III B, Pedersen T, Rosenthal Y, Seitzinger S,
Smetacek V, Steffen W. The global carbon cycle: a test of our knowledge of earth as a system.
Science. 2000;290:2916.
Field CB, Campbell JE, Lobell DB. Biomass energy: the scale of the potential resource. Trends
Ecol Evol. 2008;23:6572.
20
629
Gomez LD, Steele-King CG, McQueen-Mason SJ. Sustainable liquid biofuels from biomass: the
writings on the walls. New Phytologist. 2008;178:47385.
Horne R, Grant T, Verghese K. Life cycle assessment: principles, practice and prospects.
Collingwood: CSIRO; 2009.
Pimentel D. Global economic and environmental aspects of biofuels. Boca Raton: CRC Press;
2012.
Rosenberg NJ. A biomass future for the North American great plains: toward sustainable land use
and mitigation of greenhouse warming, Advances in global change research. New York:
Springer; 2007.
Tilman D, Socolow R, Foley JA, Hill J, Larson E, Lynd L, Pacala S, Reilly J, Searchinger T,
Somerville C, Williams R. Beneficial biofuels the food, energy, and environment trilemma.
Science. 2009;325:2701.
21
Jason G. Hamilton
Contents
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Sustainability: From Word to Concept . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Developing the Concept of Sustainability: What Is Being Sustained? . . . . . . . . . . . . . . . . . . . . . . . .
Defining the Concept of Sustainability: Focusing on Positive Change for All . . . . . . . . . . . . . . .
The Foundational Premises of Sustainability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Operationalizing Sustainability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
The Development of Sustainability Science . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Focusing Where Knowledge Is Most Needed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Sustainability Science Represents a New Conceptual Model for the World . . . . . . . . . . . . . . . . .
Future Directions in Sustainability Science . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
632
633
633
635
635
638
643
644
647
650
653
Abstract
J.G. Hamilton
Department of Environmental Studies and Sciences, Ithaca College, Ithaca, NY, USA
e-mail: jhamilton@ithaca.edu; jasonghamilton@gmail.com
# Springer Science+Business Media New York 2014
R.K. Monson (ed.), Ecology and the Environment, The Plant Sciences 8,
DOI 10.1007/978-1-4614-7501-9_18
631
632
J.G. Hamilton
Introduction
The biological sciences are traditionally organized by scale (cells, tissues, organisms, populations, communities, ecosystems, etc.), taxonomic grouping (plants,
animals, fungi, etc.), or process (competition, mutualism, evolution, etc.). In all
of these organizational schemes, there has been a tendency to view the natural
world as divided into two fundamentally different parts: humans and everything
else. Even integrative fields such as ecology have often focused on pristine
systems in the sense of trying to understand how ecosystems operate in the absence
of human influence. Although the technique of using simplified systems to understand the fundamental properties of more complex systems has a long tradition in
science (e.g., the idea of the frictionless plane developed by Galileo), the very act of
simplifying the system alters the balancing and reinforcing feedback loops that
have the potential to amplify or mute fundamental interactions between human
actions and ecosystem processes. Thus, in simplifying our view down to a system of
humans and everything else, the capacity to understand, predict, and manage the
emergent properties of our social-ecological system is lost.
In the plant sciences, the human/nonhuman dichotomy has tended to manifest
along the lines of basic versus applied perspectives. However, new fields such as
agroecology and sustainable ecosystem management are bridging this conceptual
divide. Sustainability science offers another approach that not only bridges this divide
but also explicitly connects the study of ecology (at all scales and of all taxonomic
groups) with other fields of study, especially those in the social sciences. In sustainability science, the human/nonhuman, basic/applied dichotomies are abandoned for a
21
633
new way of viewing the natural world one in which human demands on global
ecosystems is integrated into the capacity of those ecosystems to persist.
While terms such as sustainability, sustainable development, sustainability
science, and ecological sustainability are increasingly being used in both lay and
scientific vernacular, there is still much confusion regarding the meaning and
ultimately the application of these concepts. The goal of this chapter is to provide
the background and context to understand the concept of sustainability and the
relationship among sustainability, sustainability science, and ecology. In addition,
it will explore the historical development of sustainability science, provide illustrative examples of the application in sustainability science, and explore future
directions in the development of this new field.
634
J.G. Hamilton
(e.g., fitness in the Darwinian sense). There is generally no confusion for the
practitioner using the word in the new sense because the meaning is gathered from its
context. This means that knowledge about the context is required for understanding
meaning. The concept of sustainability certainly carries with it the original meaning
of the property of being sustainable with sustainable and sustain being used in
the sense of being maintained or prolonged. But, clearly, there must be something that
is being sustained or has the property of being sustainable. Thus, the defining question
becomes, what exactly is being sustained? To answer this question, it is necessary to
go back to the context under which the concept was originally developed.
The groundwork for the concept of sustainability was developed over the two
decades spanning the late 1960s to the late 1980s in a series of United Nations
reports, resolutions, conferences, and commissions. This work culminated in the
first definition and description of sustainability, articulated in the 1987 Report of the
World Commission on Environment and Development (entitled Our Common
Future and often referred to as the Brundtland Report after the chairman of
the commission, Gro Harlem Brundtland). The Brundtland Commission was
formed at a time when there was increasing recognition of and concern over the
linkages among accelerating environmental degradation, loss of natural resources,
and deterioration of peoples economic and social conditions. The members of the
commission were given a very clear charge: to propose ways to deal with environmental concerns that took into account the interrelationships between people,
resources, environment, and development. The task of the commission was ambitious it was charged with nothing less than formulating a global agenda for
change (UNWCED 1987).
The Brundtland Commission, in formulating the new paradigm for improving
overall human well-being by considering the coupled social-ecological system,
used the concept of sustainable development to create the integrating framework
of their approach. The term development was used in the broad sense of meeting
the basic needs of all people and extending the opportunity to satisfy aspirations for
a better life to everybody, with change being required in all countries, rich and poor
alike. Thus, the what of sustainability, the thing that is being maintained, is
improvement in the human condition. The report emphasizes that the sustainability
of development or sustainable development is never a fixed endpoint. Rather, it is a
process of change in which natural resource use, monetary investment, the orientation of technological development, and institutional change are consistent with
future and present needs.
While intellectually revolutionary and forward-looking in most respects, the
Brundtland Report didnt yet take the full step of recognizing the inherent systems
problem in maintaining a dichotomy between humans and the rest of the natural
world. It argued very persuasively that human well-being depends on the delivery
of goods and services supplied by well-functioning ecosystems and that ecosystem
function relies on interactions among all the component species. However, it still
described ecosystem health as a means for supporting improvements human wellbeing instead of recognizing that these two are inherently exactly the same thing.
Humans are just one component of the social-ecological systems that is the life
21
635
support system of the planet. And while the focus on human well-being as a metric
of particular concern may be chosen, it is not a distinct element from the functioning of the whole system.
Because sustainability is concerned with meeting the essential needs of the large
numbers of people on this planet whose needs are not being met, it is therefore
about creating the conditions for all people to have the opportunity to satisfy their
aspirations for a better life. Creation of these conditions involves consideration of
the functioning whole social-ecological system in which the relationships among
economy, environment, politics, and social factors are linked into a complex,
coupled system, in which no part can be viewed in isolation from the rest.
Disruption in the flows of matter and energy in natural ecosystems inevitably
leads to disruption in the flows of goods and services to humanity, thus degrading
the mean human condition. However, the negative effects of our actions are not
shared equally, thereby enabling opportunities for certain portions of humanity to
move further above the mean while others drop further below the mean. In other
words, the mean human condition has been progressively degraded and, at the same
time, that the variance around that mean has increased.
636
J.G. Hamilton
so many different contexts, that many people are beginning to feel that they must be thinking
in some wrong way about how the world works. I believe this suspicion is probably correct.
C.H. Waddington, Tools for Thought, 1977, p. xi
Why is it that problems such as global climate change, long-lived organic toxins
in our food chains, pernicious extreme poverty and hunger, lack of access to
primary education, gaps in gender equality, childhood mortality, and deadly diseases such as HIV/AIDS and malaria are proving so remarkably resistant to our best
efforts to understand and solve them? When traditional tools and approaches are not
working, progress requires a new intellectual context. The novel insight provided
by the concept of sustainability is that humans and their local and global environments exist as complex social-ecological systems. To really understand this statement, it is important to contrast the concept of complexity with the concept of
complicated. Complicated systems are just simple systems with many parts. In
simple systems, whether the parts are many or few, interactions among parts are
well defined and predictable, and thus the system is, at least in theory, well defined
and predictable. This does not mean that understanding complicated systems or the
problems arising from them is easy. For example, cars, photocopiers, and spacecraft
are complicated systems and most of us have only a tenuous grasp of how they
actually work (or how to use them)!
Complex systems consist of few or many parts, but the source and essence of
complexity arises from the richness, intensity, and character of the interactions
among constituent parts. Typically, these interactions lead to nonlinear and/or
emergent behavior (behavior that cant be predicted by studying the parts of the
system individually). Furthermore, the interactions (as well as the specific connections over which these interactions occur) constantly change, compounding the
difficulty of thorough analysis by the formation/dissolution of amplifying/stabilizing feedback loops. For example, human-induced climate change is a result of
perturbing a complex system, and finding solutions is difficult because predicting
the result of any decision strongly hinges on a thorough understanding of the
countless interactions between ecological and human social and political factors.
While definitions of sustainability can be instructive and beneficial for communication, it is impossible for any definition to convey the richness and nuance that is
being implied. In order to apply and further develop the concept of sustainability, a
deeper understanding than just a definition is required. This necessitates an understanding of the mental model on which the concept of sustainability is based.
Making the model explicit allows clear analysis of the strengths and weaknesses
of the concept and allows for implementation and improvement. One way to
succinctly describe the conceptual model of sustainability is to state it as a series
of four foundational premises. Explicit articulation of the premises can then serve
as a basis for developing research agendas, funding priorities, and mutually agreedupon courses of action.
Premise #1: The current state of human existence is not an acceptable
endpoint of societal development. Not designed to be inflammatory or accusatory,
this statement is a simple recognition that regarding the state of humanity as a
whole, we can always do better. It is not an indictment of the decisions we have
21
637
made in the past or of our current lifestyles. It is instead the fundamental driving
force that keeps us working to improve the lives of all people worldwide. While
humans have made phenomenal advances in medicine, food production and distribution, resource extraction, etc., the benefits of these advances are not enjoyed by
large portions of humanity.
Premise #2: Humans have reached a state where we are negatively
impacting the ability of future generations to meet their needs and aspirations.
The data are unequivocal that issues such as global climate change, ozone destruction, degradation of ecosystem services, depleted and limited fossil fuel resources,
accumulation of persistent toxins in the environment, new and emerging diseases,
and trends in food production all point to the same conclusion: Human impacts on
global ecosystems are accumulating at a rate that endangers our present and future
well-being. The most extensive scientific review of the data to date, contributed to
by more than 2,000 authors and reviewers worldwide, concludes:
Human activity is putting such strain on the natural functions of Earth that the ability of the
planets ecosystems to sustain future generations can no longer be taken for granted. The
provision of food, fresh water, energy, and materials to a growing population has come at
considerable cost to the complex systems of plants, animals, and biological processes that
make the planet habitable. . .. Nearly two thirds of the services provided by nature to
humankind are found to be in decline worldwide. In effect, the benefits reaped from our
engineering of the planet have been achieved by running down natural capital assets. In
many cases, it is literally a matter of living on borrowed time. (MEA 2005a)
638
J.G. Hamilton
Operationalizing Sustainability
The concept of sustainability is meant to be applied to the real world; it is a
framework for making decisions to solve problems. Sustainability has been
operationalized in a number of ways, the most common of which is to divide the
problems facing humanity into three groups: Environmental, Social, and Economic.
This division leads to the commonly used Venn-type diagrams where sustainability
is viewed as the overlap of these three realms, lenses, pillars, dimensions,
or legs (Fig. 1). This formulation has been quite attractive as it mirrors much of
our past thinking and the societal structures that have already emerged from that
thinking.
For example, this model maps easily onto existing academic disciplines and
university departmental/school structures (natural sciences, social sciences, and
economics/business), with the associated funding streams and research programs.
It also mirrors the way governmental agencies are set up in many countries. In the
United States, for instance, there is the Department of the Interior, the Department
of Commerce, and the Department of Health and Human Services. This model also
maps well onto existing NGOs and special interest groups, such as environmental
groups, social justice groups, and free trade advocates.
While relatively easy to apply and useful in some respects, this model of viewing
and operationalizing sustainability has resulted in much of the controversy, confusion, and misapplication surrounding the concept. The reductionist approach of
dividing sustainability into parts is in direct opposition to the interdisciplinary
systems approach that has propelled the concept of sustainability to its current
positions as a fundamental organizing principle for a modern form of ecology and
paradigm for future global development. One of the benefits of focusing on the
integrated premises of sustainability is that the deficiencies of the realms
approach are immediately illuminated: (1) In the real world, there arent different
realms of problems facing humans (e.g., climate change cannot be confronted
through isolated social, economic, or environmental approaches). (2) Defining a
set of realms invites focus on the artificial boundaries that differentiate the realms
instead of the whole system. This is the intellectual equivalent of the well-known
mistake of dividing a complex system into an arbitrary set of subsystems and
21
639
attempting to understand the whole system by just studying the subsystems. The
complex coupled social-ecological system is just that a system that must be
studied as a single articulated system. (3) Existing institutional (and thought)
structures must be fundamentally altered the structure of the existing systems
themselves has introduced unintended reinforcing feedbacks that allow the systems
to persist and resist adaptive change in humanecosystem interactions. The concept
of sustainability is truly transdisciplinary in nature and requires a rethinking of all
traditional boundaries. The Venn-type diagram encourages a mental disaggregation
of sustainability into parts that can be assimilated into existing intellectual frameworks, but that further entrench existing social-ecological paradigms. The result of
this can be seen in common use of terms such as ecological sustainability, social
sustainability, and economic sustainability. Because the word sustainability
refers to sustainable (as an ongoing) improvement in the state of the socialecological system, these terms are either nonsensical or congruent.
Another way to operationalize sustainability is to formulate a true systems model
that captures much more of the complexity of social-ecological systems. For
example, a model of the earth system (see Fig. 2) can be coupled with a model of
social processes (see Figs. 2 and 3).
Use of this type of model provides focus to whole-system processes and is
helpful in studying fundamental system structure and process, but they are often
too complicated to serve as tools for accurate prognosis or problem-solving. Also, it
is very difficult to use a process model to make decisions regarding time-dependent
resource allocation without extensive computer simulation. What is needed is a
simple model that captures enough of the complexity of the real situation, can be
used as a dashboard to measure progress toward goals, is evocative for thinking
about connections and possibilities, and is analytical in its mathematical structure.
One promising alternative that is being used increasingly by both scientists and
policy makers is the rose diagram (sometimes called an orientor star). A rose
diagram is a pie chart variant with each sector of the circle having the same size.
640
J.G. Hamilton
Fig. 2 System model of ecological part of the coupled social-ecological system (From Mooney
et al. (2013))
Variables are plotted as distance from the center either on the radial lines of the
sectors or by filling in the sectors themselves. Concentric circles radiating from the
center can be used for semiquantitative or even quantitative rendering. Rose diagrams are excellent conceptual ways to generate immediate visual representation of
a large number of variables (see Fig. 4).
An excellent example of how these types of diagrams are being used to organize
thinking, inform decisions, and operationalize sustainability is the United Nations
Global Compact Cities Program (see Fig. 5). In this case it is possible to consider
28 variables simultaneously. By grouping variables in appropriate ways, this
dashboard can show where efforts are having their greatest successes and
where more effort and resources should be applied. In this example, the variables
representing the social part of the social-ecological system (economics, politics,
and culture) are roughly in similar states of acceptability, while those representing
the ecological part (ecology) are, in some cases, reaching critical levels.
A weakness of the rose diagram conceptual approach is that it lacks a focus on
social-ecological processes; it doesnt show relationships among parts. It cant be
used to predict how various feedbacks among dynamic variables operate, and it
doesnt predict how leakage from one area to another might occur (i.e., how
improvement in one area might cause decline in another). At the same time, it is
extremely useful for getting a quick snapshot of a large number of important
considerations. It allows for easy expansion in numbers of variables and in numbers
of groupings of variables. Further, it avoids the Siren call of reductionism inherent
Fig. 3 System model of the social part of the social-ecological system (From Mooney et al. (2013))
21
641
Fig. 4 Example of rose diagrams used to compare trade-offs in land use. In this case, the state of each variable of interest is shown along each axis
(Reproduced from Foley et al. (2005))
642
J.G. Hamilton
21
643
Fig. 5 Example of a rose diagram that allows planners to view 28 variables simultaneously as a
sustainability dashboard of the current state of a system. (From http://citiesprogramme.com/
aboutus/our-approach/circles-of-sustainability)
in the Venn-type diagrams; for practical reasons, scientists and planners may still
focus primarily on one quadrant of the whole, but it is always clear that each
quadrant is an inseparable part of the whole. If the system starts to heavily favor one
set of factors over another, it is immediately apparent.
644
J.G. Hamilton
has at its core the same philosophy as Lubchenco in her call for use-inspired basic
research. The field explicitly emphasizes research on the fundamental character of
interactions of the social-ecological system, as well as application of this knowledge to advance sustainability goals relevant to water, food, energy, health, ecosystem services, etc. It takes the traditional focus of ecology into a new realm of
research away from the study of pristine ecosystems isolated from anthropogenic
influences and toward the study of humans as a dominating force causing change in
ecosystem states and processes. Since its first description in NRC (1999), sustainability science has become an accepted discipline in its own right, with several
specialized journals and an approximately exponential increase in numbers of
publications worldwide (Fig. 7).
21
645
Fig. 6 The relationship among traditional basic and applied research and use-inspired research
based upon the work of Donald Stokes (Clark 2007). Basic research is epitomized by the work of
Bohr to determine atomic structure; applied research is epitomized by Edison in his work to
commercialize electric lighting; use-inspired basic research is epitomized by the work of Pasteur.
Work that explores particular phenomena without consideration for generality or application is
represented in the upper left quadrant (Pastures Quadrant, Donald Stokes 1997)
Fig. 7 The temporal evolution of sustainability science as depicted by the number of publications
per year (From Bettencourt and Kaur (2011))
desire, politics, morality, etc. Probably, it is no coincidence that at the same time
that sustainability science was germinating as a discipline, the global community
was unifying to determine the worlds most urgent sustainability challenges. In
September 2000 the UN General Assembly adopted the United Nations Millennium
Declaration laying out what has now become known as the Millennium Development Goals. These goals have since served as a framework for designing the
organizing principles of sustainability science. While the terminology surrounding
what is now known as the social-ecological system of sustainability science had not
yet been developed, it is clear that the Millennium Development Goals were
oriented toward focusing policy makers and scientists toward a systems approach
to sustainability. These primary goals are generally listed as follows:
646
J.G. Hamilton
Fig. 8 Organizing research agenda of sustainability science around spatial scale and economic
status (From Kates et al. (2001))
1.
2.
3.
4.
5.
6.
7.
21
647
(v) Can scientifically meaningful limits be defined that would provide effective
warning for instabilities or tipping points in humanecosystem interactions?
(vi) How can society most effectively guide or manage humanecosystem interactions toward a sustainability transition, reversing degradation in the condition of both human societies and natural ecosystems?
(vii) How can the sustainability of alternative pathways of environment and
development be evaluated?
648
J.G. Hamilton
Fig. 9 Ecosystem services and their linkages to human well-being (From MEA (2005b))
21
649
Fig. 10 Conceptual model showing interaction among biodiversity, ecosystem services, human
well-being, and drivers of change. Changes in indirect drivers that affect ecosystem function can
lead to changes in direct drivers of ecosystem function. These changes affect ecosystem services,
which, in turn, affect human well-being (From the MEA (2003))
650
J.G. Hamilton
Fig. 11 The analytical approach of the MA that serves as a model for sustainability science in
general (From MEA (2003))
21
651
652
J.G. Hamilton
21
653
References
Beddoe R, Costanza R, Farley J, Garza E, Kent J, Kubiszewski I, Martinez L, McCowen T,
Murphy K, Myers N, Ogden Z, Stapleton K, Woodward J. Overcoming systemic roadblocks to
sustainability: the evolutionary redesign of worldviews, institutions, and technologies. Proc
Natl Acad Sci U S A. 2009;106(8):24839.
Berwick DM. A primer on leading the improvement of systems. Br Med J. 1996;312
(7031):61922. Available at http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2350403/
Bettencourt LMA, Kaurc J. Evolution and structure of sustainability science. Proc Natl Acad Sci U
S A. 2011;108(49):195405.
Carpenter SR, Mooney HA, Agard J, Capistrano D, DeFries RS, Daz S, Dietz T, Duraiappah AK,
Oteng-Yeboah A, Pereira HM, Perrings C, Reid WV, Sarukhan J, Scholes RJ, Whyte
A. Science for managing ecosystem services: beyond the Millennium Ecosystem Assessment.
Proc Natl Acad Sci U S A. 2009;106:130512.
Clark WC. Sustainability science: a room of its own. Proc Natl Acad Sci U S A. 2007;104:17378.
Daily GC, editor. Natures services: societal dependence on natural ecosystems. Washington, DC:
Island Press; 1997.
Foley JA, DeFries R, Asner GP, Barford C, Bonan G, Carpenter SR, Chapin FS, Coe MT, Daily
GC, Gibbs HK, Helkowski JH, Holloway T, Howard EA, Kucharik CJ, Monfreda C, Patz JA,
Prentice IC, Ramankutty N, Snyder PK. Global consequences of land use. Science. 2005;309
(5734):5704.
Kates RW. What kind of science is sustainability science? Proc Natl Acad Sci U S A. 2011;108
(49):1944950.
Kates RW, Clark WC, Corell R, Hall JM, Jaeger CC, Lowe I, McCarthy JJ, Schellnhuber HJ,
Bolin B, Dickson NM, Faucheux S, Gallopin GC, Gr
ubler A, Huntley B, Jager J, Jodha NS,
Kasperson RE, Mabogunje A, Matson P, Mooney H, Moore III B, ORiordan T, Svedin
U. Sustainability science. Science. 2001;292(5517):64142.
Lubchenco J. Entering the century of the environment: a new social contract for science. Science.
1998;279:4917.
MEA: Millennium Ecosystem Assessment. Ecosystems and human well-being: a framework for
assessment. Available at http://www.unep.org/maweb/en/Framework.aspx (2003).
MEA: Millennium Ecosystem Assessment. Living beyond our means, natural assets and human
wellbeing, statement from the board. Available at http://www.maweb.org/documents/docu
ment.429.aspx.pdf (2005a).
MEA: Millennium Ecosystem Assessment. Ecosystems and human well-being: scenarios, vol
2. Available at http://www.unep.org/maweb/en/Scenarios.aspx (2005b).
MEA: Millennium Ecosystem Assessment. Ecosystems and human well-being: multiscale assessments: findings of the Sub-global Assessments Working Group. Island Press. Available at
http://www.unep.org/maweb/en/Multiscale.aspx (2005c).
Meadows DH. Thinking in systems. White River Junction: Chelsea Green Publishing; 2008.
Mooney HA, Duraiappah A, Larigauderie A. Evolution of natural and social science interactions
in global change research programs. Proc Natl Acad Sci U S A. 2013;110:366572.
NRC: National Research Council Board on Sustainable Development. Our common journey: a
transition toward sustainability. Washington, DC: National Academy Press; 1999.
Reid W, Berkes F, Wilbanks T, Capistrano D. Bridging scales and knowledge systems: concepts
and applications in ecosystem assessment. Washington, DC: World Resources Institute,
Millennium Ecosystem Assessment, Island Press; 2006.
654
J.G. Hamilton
Stokes D. Pasteurs quadrant: basic science and technological innovation. Washington, DC:
Brookings Institution; 1997.
UNESCO. UNESCO and sustainable development. http://unesdoc.unesco.org/images/0013/
001393/139369e.pdf (2005).
United Nations General Assembly Resolution 42/187. http://www.un.org/documents/ga/res/42/
ares42-187.htm (1987). Accessed 11 Dec 1987.
UNWCED: United Nations World Commission on Environment and Development. Our common
future (Brundtland report). Oxford: Oxford University Press. http://www.un-documents.net/
our-common-future.pdf (1987).
Waddington CH. Tools for thought: how to understand and apply the latest scientific techniques of
problem solving. New York: Basic Books; 1977.
Further Reading
Blewitt J, editor. Understanding sustainable development. London: Routledge; 2008.
Gunderson LH, Holling CS, editors. Panarchy: understanding transformations in human and
natural systems. Washington, DC: Island Press; 2001.
Handl G. Declaration of the United Nations conference on the human environment. Audiovisual
Library of International Law. http://untreaty.un.org/cod/avl/ha/dunche/dunche.html
IPCC. Climate change 2007: synthesis report. Contribution of Working Groups I, II and III to the
fourth assessment report of the Intergovernmental Panel on Climate Change. In: Core Writing
Team, Pachauri RK, Reisinger A, editors. Geneva: IPCC; 2007. 104 pp. Available at http://
www.ipcc.ch/publications_and_data/ar4/syr/en/contents.html (2007).
Lynam T, Brown K. Mental models in humanenvironment interactions: theory, policy implications, and methodological explorations. Ecol Soc. 2012;17(3):24.
National Institute of Allergy and Infectious Diseases. Emerging and re-emerging infectious
diseases. Available at http://www.niaid.nih.gov/topics/emerging/Pages/Default.aspx
Norberg J, Cumming GS, editors. Complexity theory for a sustainable future. New York: Columbia University Press; 2008.
Staudinger MD, Grimm NB, Staudt A, Carter SL, Chapin FS III, Kareiva P, et al. Impacts of
climate change on biodiversity, ecosystems, and ecosystem services: technical input to the
2013 National Climate Assessment. Cooperative Report to the 2013 National Climate Assessment. 2012. 296 p. Available at http://assessment.globalchange.gov
United Nations. We can end poverty 2015, millennium development goals. Background. Available
at http://www.un.org/millenniumgoals/bkgd.shtml
Volk T. Gaias body: toward a physiology of earth. Cambridge: MIT Press; 2003.
Walker B, Salt D. Resilience thinking: sustaining ecosystems and people in a changing world.
Washington, DC: Island Press; 2006.
Waltner-Toews D, Kay JJ, Lister NE. The ecosystem approach: complexity, uncertainty, and
managing for sustainability. New York: Columbia University Press; 2008.
Index
A
Abiotic filtering, 7374
Aboveground net primary production
(ANPP), 212
Abscisic acid (ABA), 12
Absorbed photosynthetically-active radiation
(APAR), 209210, 215, 225
Actinorhizal N-fixing, 182
Actinorhizal symbioses, 183
Actual evapotranspiration (AET), 218
Aerenchyma, 429430
Alkaloids, 155
Allometry, 495
Alpine plants
carbon and nitrogen storage, 345348
convection, 343
CO2 availability, 349352
description, 330
latent heat exchange and water, 342
microclimate and energy balance, 335338
physiological and ecological constraints,
353
radiation stress, 351353
radiation, 339342
soils, 334335
stress response and growth strategies, 345
temperature stress, 348349
temperature, 338339
Analytical flow cytometry (AFC), 488
Anemophily, 92
Anoxia, 429
Anthropocene, 534
Ants transport, 104
AOT40 595
Araucaria cunninghamii, 34
Arbuscular mycorrhizae (AM), 185
Arbutoid mycorrhizae, 187
Arms race, 146
655
656
Cavitation, 313
Censusing populations, 6163
Chiropterophily, 92
Climate change, 408
Climate, 593594
Coastal squeeze, 451
Coherent wind structures, 586
Community assembly, 6871, 84
Complex adaptive systems (CAS), 651
Convection, 343
Corpse flower, 90
CO2 fertilization effect, 8
C4 photosynthesis, 402
Cyanogenic glycosides, 154
D
Damkohler number, 584
Demographic parameters, 33
Demographic stochasticity, 47
Denitrification, 447
Density-independent limiting forces, 41
Desert, 367
Detritus, 229
Developmental noise, 122
Diaheliotropism, 307
Dioecious plants, 96
Direct inducible defenses, 169
Dispersal, 7173
Drought, 302, 407
Dwarf mangrove habitat, 443
E
Ecological succession, 278280
Ecosystem respiration (ER), 211
Ecosystem services, 543
Ectendomycorrhizae, 187
Ectomycorrhizae (EM), 185187
Edaphic factors, 221
Emission factor, 578, 581
Enemy release hypothesis, 46
Energy loss mechanisms, 286
Environmental stochasticity, 46
Environmental variability, 55
Enzymatic hydrolysis, 608
Epiphyte grazers, 469470
Ericoid mycorrhizae, 187
Erosion control, 448449
Escape efficiency, 587
Estuary, 427
Eutrophication, 452
Evolutionary arms race, 147
Index
F
Feeding deterrents, 162163
Fertilizer, 621
Fire, 289290, 408412
Flora, Arctic, 369
Flux measurements, 19
Fog deserts, 301
Food crops, 612
Forest gap models, 279
Forest, 275
Functional equilibrium hypothesis, 237
Functional traits, 70
G
Gasification, 609
Genetic diversity, 477479
Geometric model, 40
Geomorphic processes, 382385
Geukensia demissa, 435
Glucosinolates, 154155
Grazing, 412416
Greenhouse effect, 535536
Greenhouse gas emissions, 537539,
617618
Gross primary production (GPP), 209, 210,
505508
H
Habitat filtering, 73, 79
Haleakala silversword, 30
Halophytes, 316, 427
Harmful algal blooms (HABs), 524
Heat shock proteins, 555
Heliotropism, 306
Hermaphroditic flowers, 96
Heterocystous cyanobacteria, 184
Human activities, 112114
Humanity, 636
Human/non-human dichotomy, 632
Hydraulic lift, 312
Hydric soils, 426
Hydrodynamics process, 467468
Hydrolysis, 608
Hypoxia, 427
I
Indirect inducible defenses, 170171
Indirect land-use change (iLUC), 616
Initiation reactions, 589
Insect enzymatic detoxification system, 163
Index
Interspecific competition, 322
Islands of fertility, 317
K
Krummholz, 333, 357
L
Land-use change (LUC), 616
Latent heat exchange, 337
Leaf area index (LAI), 6, 18, 210, 215,
225, 229
Leaf endophytes, 188189
Leaf energy balance, 305310
Life table, 5051
Life-cycle diagram, 52
Light-use efficiency (LUE), 215
Linanthus parryae, 42
Liquefaction, 610
Liquid biofuels, 615
Logging, 293294
Long Term Ecological Research (LTER), 216
Long wave radiation, 340341
M
Macroevolutionary hypotheses, 146147
Mainland-island model, 61
Mangrove swamps, 437
Marine protected areas (MPA), 480
Matrix model, 5455
Mean annual temperature (MAT), 240
Melittophily, 92
Meristem limitation, 228
Meta-analysis, 240
Metapopulation, 60
Michaelis-Menten model, 350
Microbial diversity
phyllosphere, 196
rhizosphere, 198
small-subunit ribosomal RNA, 196
Microclimate, 12
Microflagellates, 491
Microphytoplankton, 491492
Mid-continent deserts, 300
Millennium development goals, 645
Millennium Ecosystem Assessment (MA), 647
Mitigation, 476
Model-data fusion (MDF), 23
Monoecious species, 96
Monotropoid mycorrhizae, 187188
Mycorrhizal diversity, 199
657
N
N-fixing mutualisms, 180
Nanoflagellates, 491
Nanophytoplankton, 491
Natural communities, 234
Net community production (NCP), 508509
Net ecosystem production (NEP), 211, 216
Net primary production (NPP)
abiotic controls, 218226
ANPP, 212
biodiversity effects, 232235
biotic controls, 230231
BNPP, 213214
community change, 235
disturbance, 229230
herbivory, 236237
legacy effects, 227
remote sensing and modeling approaches,
214216
sequential limitation, 219
vegetation structure, 231232
Niche conservatism, 82
Nitrogen fixation, 447
Non-leaf photosynthetic structures, 310311
Non-native bioenergy crop, 625
Non-photochemical quenching, 352
Non-protein amino acids, 153154
Nurse plant/nurse-protege association, 322
Nutrient limitation, 518519
Nutrients, 380381
O
Ocean acidification (OA), 503, 523, 526
Ocean warming, 525
Oil palm, 267268
Ontogenetic drift, 127128
Optimal partitioning models, 124
Optimal partitioning theory (OPT), 130
Orchids, 187
Ornithophily, 92
Ozone (O3), 9, 563
P
Park Grass Experiment, 391
Pelagic environment, 496498
Permafrost, 377378
Pests and pathogens, 626
Phenolics, 156157
Phenotypic plasticity
adaptive response, 124
definition, 121
658
Phenotypic plasticity (cont.)
developmental noise, 122
importance in plants, 123
methodological approaches, 136138
nonadaptive/maladaptive, 126
role in evolution, 126127
techniques for evaluation, 127136
vs. developmentally programmed changes,
127
Photochemical smog, 594
Photoinhibition, 352
Photosynthesis, 304
Photosynthetically active radiation (PAR),
5, 206
Phreatophytes, 311
Phylogenetics, 78
Phytoliths, 400
Phytoplankton ecology, 513514
Phytoremediation, 447
Picophytoplankton, 489490
Plant functional traits (PFT), 287288
Plant functional types (PFTs), 493
Plant growth promoting rhizobacteria (PGPR),
190191
Plant-microbe interactions (PMI), 178
Plants
adaptation to environmental change,
566567
and ecosystem services, 543
in global carbon cycle, 541542
response to CO2 550551
response to drought, 558563
response to ozone, 563566
response to temperature, 551558
Poa annua, 42
Pollination syndromes, 93
Pollination
animals benefit, 9192
delivery system, 9293
dispersal agents, 92
evolutionary dynamics, 100101
genetic and evolutionary consequences,
98100
plant mating systems, 96
plants benefit, 91
Population dynamics, 32, 38, 59, 401
Posidonia australis, 462
Precipitation, 7, 302, 381, 539
Proteinase inhibitors, 155
Pseudovivipary, 437
Pycnocline, 501
Pyrolysis, 609
Index
R
Radiative heat exchange, 337
Rain-shadow deserts, 300
Ramsar Convention, 452
Relative humidity (rH), 7
Relaxed eddy accumulation (REA), 586
Remote sensing (RS), 19, 509
Residence time, 211
Resilience theory, 651
Respiratory needs (Rr), 213
Restoration
grassland, 418420
and recovery, 476477
wetland, 453454
Rhizobia-legume mutualism, 180182
Rhizosphere C flux (RCF), 237
Rhizosphere effect, 191192
Root endophytes, 188
Root systems, grassland, 404
Rose diagram, 639
Rubisco, 553
Runoff, 619620
S
Salt marshes, 431
Seagrass ecosystems
abiotic factors, 465
average vs. marine and terrestrial
ecosystems, 465
economic goods and services, 466467
epiphyte grazers, 469470
food webs, 470472
future threats, 472476
generas of, 464
genetic diversity, 477479
grazers, 469
hydrodynamics and resilience, 467468
nurseries for juvenile fish, 464
restoration and recovery, 476
species in, 459
Secondary metabolites, 144
Seed dispersal
agents, 103105
animals benefit, 102
evolutionary dynamics, 109
fruit characteristics, 105106
packaging, 103
patterns, 106109
plants benefit, 102
Sensible heat exchange, 337
Shortwave, solar radiation, 339
Index
Sky islands, 320, 329
Snowbed species, 359
Snow glades, 344
Socio-ecological system, 632
Soil organic carbon (SOC), 621622
Solar radiation (SR), 56
Source-sink model, 61
Spartina alterniflora, 430, 432, 434, 435
Species pool, 68, 83
Species richness, 232
Standing crop, 209
Stomata, 559
Structured models, 40
Suberin, 437
Subsidence, 450
Succulent plants, 314315
Superorganismic, 279
Sustainability, 632
Synthetic communities, 234
T
Temperate forests, 275276
Termination reactions, 589
Terpenoids, 155156
Thermal acclimation, 310
Thermokarst, 384
Tillers, 398
Total root allocation (TRA), 213214,
237238
Total soil respiration (TSR), 214, 238
Traditional growth analysis techniques, 134
Trapline, 98
Trichodesmium, 494
Tropical rain forests
ant-plant symbioses, 256
biogeography, 250252
cambial dormancy, 254
climate, 250252
659
factors, 259
future aspects, 268
lianas, 254
mycorrhizal associations, 257
physiognomic properties, 252
plant-pest interactions, 262
productivity and nutrient cycling, 263265
threats, 265267
tree fall gaps, 262
Tropospheric ozone, 588
Tundra, 366
Turnover coefficient (TC), 214, 237238
Turnover rate, 211
U
Unstructured models, 39
V
Vapor pressure deficit (VPD), 7
Vegetation structure constraint, 232
Vertebrate disperser, 104
Vivipary, 437
Volatile organic compounds (VOCs), 17
W
Water cycling, 285287
Water requirements and quality, 622623
Water-mediated dispersal, 104
Wetland hydrology, 426
Wildlife, 623624
Wind dispersal, 103
Wind speed, 7
Z
Zeldovich reaction, 8